Jukuri, open repository of the Natural Resources Institute Finland (Luke) All material supplied via Jukuri is protected by copyright and other intellectual property rights. Duplication or sale, in electronic or print form, of any part of the repository collections is prohibited. Making electronic or print copies of the material is permitted only for your own personal use or for educational purposes. For other purposes, this article may be used in accordance with the publisher’s terms. There may be differences between this version and the publisher’s version. You are advised to cite the publisher’s version. This is an electronic reprint of the original article. This reprint may differ from the original in pagination and typographic detail. Author(s): Teppo Vehanen, Tapio Sutela, Jukka Aroviita, Satu-Maaria Karjalainen, Juha Riihimäki, Aron Larsson, & Kari-Matti Vuori Title: Land use in acid sulphate soils degrades river water quality – Do the biological quality metrics respond? Year: 2022 Version: Publisher’s version Copyright: The author(s) 2022 Rights: CC BY 4.0 Rights url: https://creativecommons.org/licenses/by/4.0/ Please cite the original version: Teppo Vehanen, Tapio Sutela, Jukka Aroviita, Satu-Maaria Karjalainen, Juha Riihimäki, Aron Larsson, & Kari- Matti Vuori. (2022). Land use in acid sulphate soils degrades river water quality – Do the biological quality metrics respond? Ecological Indicators, vol. 141, article id 109085. https://doi.org/10.1016/j.ecolind.2022.109085 Ecological Indicators 141 (2022) 109085 Available online 25 June 20221470-160X/© 2022 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Original Articles Land use in acid sulphate soils degrades river water quality – Do the biological quality metrics respond? Teppo Vehanen a,*, Tapio Sutela b, Jukka Aroviita c, Satu-Maaria Karjalainen c, Juha Riihimaki d, Aron Larsson a, Kari-Matti Vuori d a Natural Resources Institute Finland, Latokartanonkaari 9, 00790 Helsinki, Finland b Natural Resources Institute Finland, Paavo Havaksen tie 3, 90570 Oulu, Finland c Finnish Environment Institute, Paavo Havaksen tie 3, 90570 Oulu, Finland d Finnish Environment Institute, Latokartanonkaari 11, 00790 Helsinki, Finland A R T I C L E I N F O Keywords: Fish assemblage Benthic invertebrates Diatoms Human impact Acidity Forestry A B S T R A C T Land use in the Acid Sulphate (AS) soils induces metal and acidity pollution of aquatic ecosystems in coastal areas worldwide. Increasing utilization of AS soils poses increasing risks for deterioration of water bodies. We studied the effects of the coverage of AS soils, together with other catchment land cover attributes, on aquatic assemblages of fish, diatoms and benthic invertebrates in 42 sites along 15 lowland rivers in Finland during three subsequent years. Low pH and increasing content of several metals in the river water were related to high amount of AS soils in the catchment. Especially increasing iron content and water color were correlated to amount of forested areas in the catchment, whereas lower water color values and higher arsenic, chromium and iron concentrations were associated with wetlands. The assemblage structure of all three biological groups was strongly spatially structured among rivers and varied less temporally. The spatial structure of fish and diatoms were strongly affected by the acidic water, whereas invertebrates were more affected by low alkalinity and increasing concentrations of organic matter and iron. Especially fish and benthic invertebrate bioassessment metrics demonstrated for AS soil induced degradation in acidity by responding to low pH and high acidity, while the response from the diatoms index was weaker. The high metal concentrations alone did not seem to add to the degradation in biometrics without further increase in acidification. Our results highlight the importance of recognizing AS soil areas in the catchment to target the mitigation effects. A holistic approach in the mitigation of the adverse effects from AS soils is needed, using several mitigation methods in the catchment, and directing main efforts and protection from human disturbance to catchment areas with the highest proportion of AS soils. Our results suggest that status assessment of AS rivers should be based on multiple biological quality elements and that their metrics could be improved for better detection of impacts from acidity and metal pressures. The effects of metals and their concentrations on aquatic assemblages should be further examined. 1. Introduction Freshwater ecosystems are threatened worldwide, and their biodi- versity is declining rapidly (Dudgeon et al., 2006; Woodward et al., 2010; Reid et al., 2019). The decline is typically a result of human- driven disturbances impairing biodiversity and ecosystem stability (Hautier et al., 2015). Catchment land use is one of the main causes for this development (Bayramoglu et al., 2020; Chen and Olden, 2020), and it is likely to accelerate when more and more marginal lands are taken into use (Popp et al., 2017; Chen et al., 2020). There are about 170 000 – 240 000 km2 acid sulphate (AS) soils in the world (Ritsema et al., 2000; Andriesse and van Mensvoort, 2002). Main areas of occurrence are in Africa, Australia, Asia and Latin America, typically in coastal and estuarine regions (Dent and Pons, 1995). AS soils are stable under anoxic conditions, but once they are oxidized, the sulfide minerals, mainly pyrites, form into sulfuric acid (Åstrom and Bjorklund 1996; Osterholm and Åstrom, 2002). Most of the AS soils In the Baltic Sea region result from the Litorina Sea stage some 7500 –4000 year ago when the salt content of the seawater was substantially higher than it is today (Tikkanen and * Corresponding author. E-mail address: teppo.vehanen@luke.fi (T. Vehanen). Contents lists available at ScienceDirect Ecological Indicators journal homepage: www.elsevier.com/locate/ecolind https://doi.org/10.1016/j.ecolind.2022.109085 Received 26 March 2022; Received in revised form 15 June 2022; Accepted 17 June 2022 Ecological Indicators 141 (2022) 109085 2 Oksanen 2002). Sulfate-reducing microbes reduced the sulphates in seawater into sulphides, and as a consequence subsurface anoxic soils contain metal sulphides, mainly in the form of pyrites (Gupta & Germida 2021). Postglacial land uplift in the Baltic Sea area and human activities bring these sulphide bearing layers in contact with oxygen. Human ac- tivities involve especially drainage of soil for agriculture or forestry, where open ditches and subsurface pipes are conducted to lower the level of groundwater. As a result of soil drainage and concurrent oxidation of metal sulphides, sulfuric acid is produced (Åstrom and Bjorklund 1996). This leads to acidification of the soil, characterized by very low pH and extensive leaching of metals (Åstrom and Bjorklund, 1996; Boman et al., 2010). This toxic mix spreads into the environment, typically during heavy rain periods (Toivonen and Osterholm, 2011; Nystrand and Osterholm, 2013). The most visible associated sudden impacts are fish kills (Cook et al., 2000; Sutela et al., 2012; Sutela and Vehanen, 2017), but it is obvious that there are spatially and temporally variable impacts on the whole aquatic ecosystem (e.g., Corfield, 2000; Faltmarsch et al., 2008; Toivonen et al., 2020). The temporal and spatial distribution of the impacts are, however, poorly understood. The AS soil problems culminate in the coastal and estuarine areas with intensive land use. This is true also for Finland, where the largest AS soil areas are located along the western and south-western coast (Ministry of Agriculture and Forestry and Ministry of the Environment, 2011). The total area of AS soils is estimated to be between 1000 and 3000 km2, making Finland the country with most AS soils in Europe (Yli- Halla et al., 1999; Faltmarsch et al., 2008). These areas are well suited for agriculture, and a large proportion of the AS soils are under cultivation. The European Water Framework Directive (WFD) has established indices based upon the Biological Quality Elements (BQEs) composition (e.g., fish, benthic diatoms, and macroinvertebrates) to assess ecological quality of freshwaters. The outcome of these indices has been inter- calibrated among EU countries, i.e. within the quality element they indicate similarly results of the level of disturbance on the aquatic communities (Poikane et al., 2014). The WFD assessment methods measure biological degradation from undisturbed reference conditions. For the water bodies in Finland, the same type-specific reference con- ditions are used irrespective of AS soil cover in their catchment. The expectation is that the pressures from AS soils can be suppressed to the extent for the water bodies to reach good ecological status. The WFD methods are commonly a combination of multiple metrics aimed at reflecting the impact of most common pressures, like eutrophication and hydro-morphological degradation (Poikane et al., 2020). Only three countries have fish and benthic invertebrate methods targeted to detect the effects of acidification in WFD classification (Sweden, Norway, and UK; Poikane et al., 2020). There are, however, also other pressure- specific methods developed to detect the effect of acidity on biota in rivers (e.g. Juggins et al. 2016). Given the importance of the assessment outcomes to guide the necessary actions needed to mitigate the negative effects, it is important that they show the connection between the sources of disturbance and the true ecological conditions (Rapport and Hilden, 2013). Aquatic community structures of fish, invertebrates and diatoms often show concordance with respect to environmental gradients (Jackson and Harvey, 1993; Kilgour and Barton, 1999; Soininen et al., 2009). In rivers these environmental gradients and aquatic communities are typically longitudinally arranged along the river continuum from headwaters to estuaries (Rosi-Marshall and Wallace, 2002; Cross et al., 2013; Sutela et al., 2020). However, the community concordance in not always consistent, as the structure of aquatic communities is a result of biogeographical processes acting at multiple habitat scales (Allen et al., 1999; Bae et al., 2011; Vehanen et al., 2020). Further, several other factors like dispersal and functional traits affect the assemblage composition (Grenouillet et al. 2008; Feio et al., 2017) and the concordance among aquatic communities in their response to distur- bance can be weak (Carlisle et al., 2008; Pace et al., 2012). Therefore, the use of more than one biological quality element for the ecological impact assessments is recommended (Pace et al., 2012; Bae et al., 2014; de Morais et al., 2018). Here we study 1) how the proportion of AS soils of the catchment area affects river water quality, 2) how the water quality impacts depend on the interactions between AS soils, land use and catchment charac- teristics 3) if the proportion of AS soils in the catchment affects spatio- temporal variation of the fish, benthic invertebrate and diatom assem- blage structures, and 4) what is response of the national WFD indices of the biological quality elements to the disturbance from AS soils. 2. Material and methods We sampled 15 lowland streams flowing to Baltic Sea in western and south-western coast of Finland during three subsequent years 2010–2012 (Fig. 1, Table 1). Mean flow of the streams ranged from 0.9 to 48.8 m3 s 1 during the study period, thus representing small to mid- sized streams (Table 1). From each river 2–6 swiftly flowing rapid reaches were sampled for water chemistry and biological communities (Fig. 1, Table 1). Landcover attributes and land use in the catchment area above each reach were determined using GIS analyses. Landcover attributes included agricul- tural area, forested areas, wetlands, area of surface waters (Corine Land Cover 2012) and the area of AS soils (Table 2). The area of AS soils was obtained from the digital data of the Geological Survey of Finland (GTK). GTK has created an interpretation map of the occurrence of acid sulfate soils in Finnish coastal areas up to the highest shoreline of the ancient Litorina Sea (https://gtkdata.gtk.fi/hasu/index.html). The interpretation is created using quantitative multivariable modelling based on spatial analysis. Interpretation map was used because all catchments above our study rapids were not physically sampled for AS soils (out of total of 42 rapids, 14 were not sampled from catchment above the rapid). We could, however, confirm the reliability of the interpretation by correlating the areas with sampling data with the interpretation data (Pearson correlation coefficient r ˆ 0.959, N ˆ 28, p < 0.001). The degree of land use pressure from agriculture and urban- ization in each catchment was estimated by the area used for agriculture and the artificial surfaces (Table 2). All variables were calculated as proportion of the total area of the catchment above the reach (per km2). The land use pressure from forestry was estimated by calculating the total length of ditches (km) in the catchment, i.e., drainage intensity per catchment area (Topographic database, National Land Survey of Finland) (Table 2). Water quality samples were analyzed using standard methods by the Finnish Environmental Institute (https://www.syke.fi/methodsstanda rdization) and the data are stored in the open national database Hertta (https://wwwp2.ymparisto.fi/scripts/kirjaudu.asp). Yearly mean value was used for each water quality variable (mean number of samples per year and rapid ˆ 6, median ˆ 4, range 2–31). Water quality variables included metals (µg/l) (aluminum (Al), arsenic (As), cadmium (Cd), cobalt (Co), chromium (Cr), copper (Cu), lead (Pb), sodium (Na), nickel (Ni), iron (Fe) and zinc (Zn)), conductivity, pH, acidity (mmol/l), turbidity (FNU) and color (mg/l Pt) (Appendix 1). Environmental Quality Standards (EQS) for metal content in surface waters are listed for Cd, Pb and Ni in EU Directive 2008/105/EC. The directive listed priority substances for achieving good chemical status for water bodies in Europe. For the remaining metals, excluding Na, Finland recommends using Canadian Freshwater Metals Criteria (https://www.flowlink.ca /freshwater-metals-criteria). Fish assemblages were sampled in each reach each year in August- September by electrofishing following the sampling procedure for WFD monitoring (Vehanen et al., 2010) Fish were captured using Hans Grassl Gmbh (Schonau am Konigssee, Germany) 1G 200–2 electro- fishing gear using pulsed (50 Hz) DC current with 400–600 V voltage. All captured fish were identified to species and counted. Total length (TL) of the narcotized fish was measured to the nearest 1 mm and individuals of T. Vehanen et al. Ecological Indicators 141 (2022) 109085 3 each species were weighed to the nearest 0.1 g. Age-0 ‡ brown trout, Salmo trutta, were recorded separately from older age classes. After re- covery from anesthesia, the fish were released back to the stream. Benthic diatoms were sampled using methods in accordance with the European standard (SFS-EN13946 2005) following the national sam- pling procedure for WFD monitoring (Jarvinen et al., 2020). For each site sampled diatoms were brushed with a toothbrush from the surface of five randomly collected cobbles. The results were pooled into one composite sample and preserved in ethanol at the site. In the laboratory samples were treated with strong acid solution (HNO3 ‡H2SO4; 2:1) and identified to species level with differential interference contrast (1000x magnification). Approximately 500 diatom valves were counted to determine the relative abundance of the diatom species. Benthic macroinvertebrates were sampled and processed following the national sampling procedure for WFD monitoring (Jarvinen et al., 2020). A composite 2-min kick-net sample was taken at each sampled reach using a net mesh size of 500 μm. The samples were preserved in 70% ethanol in the field. In the laboratory the samples were sorted, and all individuals were identified to the lowest possible taxonomic level, usually species or genus with the exclusion of Chironomidae (Diptera) and Oligochaeta. Identification was undertaken using national lists and keys. WFD compliant biological assessment indices and their Ecological Quality Ratios (EQR), that are used to estimate the ecological status, were calculated for fish (Vehanen et al., 2010; Pont, 2011), macro- invertebrates (Aroviita et al., 2008) and diatoms following the current Fig. 1. Location of the 15 study rivers and rapids (2–6 in each river) used to study the effects of acid sulphate soils on water quality and aquatic assemblages. T. Vehanen et al. Ecological Indicators 141 (2022) 109085 4 national guidance (Aroviita et al., 2019). The fish index (FiFI) consists of five metrics: proportion of intolerant species, proportion of tolerant species, density of age-0 ‡ brown trout and salmon, Salmo salar, density of cyprinid group individuals and number of species present. The mac- roinvertebrate assessment includes three metrics that measure compo- sitional dissimilarity to reference conditions: number of stream type -specific taxa, number of stream type -specific EPT families (mayflies (Ephemeroptera), stoneflies (Plecoptera) and caddisflies (Trichoptera)) and PMA-index (Percent Model Affinity, Novak and Bode, 1992). The diatom assessment uses two metrics: number of stream type -specific taxa and PMA-index. We calculated the mean EQR (mean  SD) for each of the sampled rivers (N ˆ 15 for fish, 12 for diatoms, and 10 for benthic Table 1 The 15 rivers studied and their characteristics. Sampled organism groups are also indicated. River Length km Catchment area km2 MQ m3/s HQ m3/s NQ m3/s Fish Diatoms Macro-invertebrates Viirretjoki 14.0 195.4 1.4 191.0 0.0 X X X Kalvianjoki 28.5 324.0 2.5 32.0 0.1 X X X Narpionjoki 76.5 991.9 9.0 135.4 0.2 X X Tiukanjoki 60.0 500.0 4.9 98.0 0.1 X X X Maalahdenjoki 27.5 677.2 4.4 52.0 0.1 X X X Harrstrom 12.0 139.8 1.1 11.1 0.1 X X X Laihianjoki 42.4 506.0 3.9 47.0 0.2 X X X Voyrinjoki 35.0 222.7 2.5 19.9 0.1 X X X Kimojoki 18.4 196.2 2.2 17.6 0.1 X X X Purmonjoki 68.6 864.3 8.7 22.4 1.0 X X Lestijoki 99.9 1373.0 14.7 137.0 2.1 X X X Kainastonjoki 23.8 424.2 3.5 30.8 0.5 X Kyronjoki 131.7 4923 48.8 409.0 3.8 X Isojoki 52.4 693.2 13.3 199.0 1.4 X Lillån 13.9 100.8 0.9 12.0 0.1 X X X Table 2 Catchment area and upstream catchment characteristics of the 42 rapids sampled in the 15 study rivers. The area of agriculture, AS soils, forests, wetlands, waters and artificial surfaces are given as percentage of the catchment area, forest ditches as kilometers per km2. River name Rapid name Catchment area km2 Agriculture % AS soils % Forests % Ditches km per km2 Wetlands % Waters % Artificial surfaces % Viirretjoki Tiinukoski1 166.8 14.9 16.4 76.6 10.3 5.9 0.1 2.5 Tiinukoski2 166.9 14.9 16.4 76.6 10.3 5.9 0.1 2.5 Alakoski 187.6 15.8 16.2 75.4 10.1 5.8 0.2 2.8 Kalvianjoki Myllykoski 121 4.8 1.2 87.1 11.9 6.7 0.7 0.7 Honkalankoski 131.4 5.5 2.2 86.8 11.6 6.3 0.7 0.7 Keskusta 268.1 7.9 6.2 83.6 10.6 6.5 0.4 1.6 Narpionjoki Stenforsen 744.9 17.9 10.5 71.0 11.0 7.1 0.8 3.2 Allmanningsf. 939.8 20.1 14.5 69.8 10.3 6.0 0.7 3.4 Backforsen 944.4 20.3 14.6 69.6 10.2 6.0 0.7 3.5 Tiukanjoki Tiukanjoki 3 22.1 11.3 0.0 84.2 17.8 3.2 0.0 1.8 Tiukanjoki 2 280.5 22.7 12.9 69.6 12.3 3.3 0.0 4.4 Puskamarkki 485.8 20.2 12.8 73.0 12.3 3.1 0.1 3.6 Maalahdenjoki Sågkvarnsfors 133.6 11.1 13.2 81.8 9.9 5.2 0.1 1.9 Kyrkbacken 427.3 14.2 19.2 79.9 9.9 3.2 0.1 2.5 Maalahdenjoki 1 492.8 15.2 20.4 78.7 9.7 3.3 0.1 2.6 Harrstrom Harrstrom 3 76.1 5.0 12.4 76.5 9.8 7.5 9.3 1.6 Harrstrom 2 76.8 4.9 12.2 76.7 9.8 7.4 9.2 1.6 Harrstrom bro 139.8 13.3 17.5 72.0 8.1 7.0 5.1 2.6 Laihianjoki Laihianjoki 3 198.6 10.4 14.1 81.3 9.7 6.2 0.3 5.2 Laihianjoki 2 298.7 17.8 20.4 74.7 7.7 4.6 0.3 2.7 Yrjaala 306.3 18.7 21.7 73.2 7.5 4.4 0.3 1.1 Voyrinjoki Voyrinjoki 3 77.3 11.9 28.5 83.7 9.2 3.1 0.1 1.2 Voyrinjoki 2 96.9 16.6 31.5 78.9 8.0 2.8 0.1 1.5 Kimojoki Kimojoki 2 90.3 19.4 5.5 65.4 4.3 7.3 5.1 2.9 8-tien silta 204.8 20.1 8.9 69.3 5.3 4.8 2.7 3.0 Purmonjoki Purmojoki 3 476.4 19.2 13.2 71.9 7.8 4.2 1.8 2.9 Purmojoki 1 817.4 16.4 9.3 74.1 7.9 4.9 2.2 2.4 Lestijoki Javajankoski 1100.6 9.9 0.1 70.5 10.0 10.7 7.4 1.5 Kuustonkoski 1291.8 10.5 2.3 71.1 9.9 10.1 6.4 1.9 Niskankoski 1300.3 10.5 2.3 71.2 9.9 10.1 6.4 1.9 Sampilankoski 1306.3 10.5 2.3 71.2 9.9 10.0 6.4 1.9 Roukalankoski 1312.1 10.5 2.4 71.3 9.9 10.0 6.4 1.9 Raumankoski 1386.9 10.7 2.8 71.7 10.0 9.6 6.0 1.9 Kainastonjoki Pohjapato 130.2 29.0 3.8 61.3 10.8 6.8 0.0 2.8 Natynkoski 413.1 29.4 1.5 60.2 9.9 6.3 0.1 3.9 Kyronjoki Perttilankoski 3896.6 22.7 6.0 65.1 8.6 6.3 1.5 4.4 Voitila 4814 24.9 12.2 63.9 7.9 5.4 1.4 4.4 Isojoki Nybrofors 1049.2 12.6 4.3 78.7 11.2 6.1 0.4 2.2 Tuimalankoski 471.1 11.4 1.8 81.3 11.8 5.0 0.3 2.0 Talvitienkoski 400 11.2 1.5 82.3 12.1 4.2 0.3 2.0 Lillån Lillån 2 97.1 11.9 4.9 81.2 12.6 4.9 0.0 1.9 Valtatie 8 97.5 11.9 5.0 81.1 12.6 4.9 0.0 1.9 T. Vehanen et al. Ecological Indicators 141 (2022) 109085 5 invertebrates) over years and study sites. For the diatoms we calculated also one pressure specific index, ACID index, especially developed in the same region (Sweden) to indicate acidic pressures (Andren & Jarlman, 2008). The index consists of two parts. First is the ratio between the abundances of the species complex around Achnanthidium minutissimum (ADMI) and Eunotia-species (EUNO) (ADMI/EUNO ratio). The second part is the entire diatom community classified circumneutral (neutr) and alkalibiontic ‡ alkaliphilous(alkbp) species, divided by acidobiontic ‡ acidophilous (acibp) diatoms (Andren & Jarlman, 2008). 2.1. Statistical analyses Multivariate analyses were used to summarize the key indicators of water quality and their variability in the study sites, and examine their relationship to catchment landcover attributes. We also analyzed how the aquatic assemblages were structured temporally (years) and spatially (rivers), and which water quality and landcover attributes affected their structure. Principal component analysis (PCA) was used to reduce the number of water quality variables into mutually uncorrelated linear combinations (Abdi and Williams, 2010). PCA on water quality variables was computed by the base R-statistics package “prcomp”, with centering so that variables are shifted to be zero centered and scaled to have unit variance before analysis. The calculation was done by a sin- gular value decomposition of the centered and scaled data matrix, not by using eigen on the covariance matrix. PCA plots were done with “fac- toextra” package (Kassambara and Mundt, 2020). Horn’s parallel anal- ysis (Horn 1965) was used to determine the number of components to keep in the analysis. Statistical analyses were made using the statistical computing software R version 3.6.0 (R Core Team, 2019). Canonical correlation analysis (CCA) explores to quantify and iden- tify the interrelationships among sets of criterion variables and predictor variables (Weenink, 2003). CCA was used to examine the associations of water quality (the three first water quality principal components) to landcover attributes. CCA was computed and plotted with the R-package “yacca” (Butts, 2018). Canonical correlations were computed with rotated data. Both X and Y variables were scaled to unit variance, scaling the output of coefficients for the variables on each canonical variate. Rao’s F approximation and Bartlett’s Chi-squared test were used to test the significance of canonical correlations. Quantities of particular in- terest included the correlations between the original landcover attri- butes variables in each set and their respective canonical variates in water quality variables (structural correlations or loadings). To explore the temporal and spatial similarity of aquatic assemblages Bray-Curtis dissimilarity indices were computed for fish, benthic invertebrate and diatom data using the vegdist function in the package “vegan” (Oksanen et al., 2019). Fish species densities (ind. per 100 m2) were used to analyze fish assemblage, relative abundance for diatoms and number of individuals per sample for benthic invertebrate species (Appendix 2). A dummy variable (1) was added to each sample in the original data prior to computing the dissimilarity indices to avoid fish samples with all zero scores. The dissimilarity indices were reduced to Principal Coordinates (PCo) using the betadisper function, which im- plements the PERMDISP2-procedure (Anderson, 2001). Permutation tests of the PCo’s were performed with the ANOVA-like function per- mutest. Principal Coordinates were used to display the assemblage structure among rivers and years in ordination space. Finally, PERMA- NOVA (Anderson, 2001) was computed using the function adonis to analyze differences between the rivers and sampling years and their interaction in species assemblages. Rivers and years were treated as fixed factors in the PERMANOVA. 3. Results There were clear gradients in the water quality of the study rivers (Appendix 1). The first three principal components of the PCA analysis explained 48.6 %, 14.7 % and 10.9 % of the variance in the water quality variables, respectively (Fig. 2). Horn’s Parallel Analysis indicated to retain these three principal components, having eigenvalues above one, in the analysis. On principal component one, correlations between the original variables and principal components showed that content of several metals (Al, Cd, Co, Cu, Na, Ni and Zn) increased with increasing acidity and conductivity increased as the concentration of metal ions increased (Fig. 2). Second component mainly illustrates alkalinity, and, to a lower extent, Pb and turbidity, while Fe and color showed a negative correlation to these water quality metrics (Fig. 2). The main variables in the third principal component were water color, As and Cr, all having moderate positive correlations (Fig. 2). In general, the average and the maximum metal contents in the river water of the study sites was higher than the EQS criteria (Table 3). The exception to this were As, Pb and Ni. In a revision of the EU Directive 2008/105/EC in 2013, the original Ni EQS (20 µg/l) has been revised to 4 µg/L bioavailable Ni. The variation in metal contents was relatively high, which could be expected given the range of AS soils (Table 2 and Table 3). Significance tests on CCA showed statistical significance indicating a strong interrelationship between the two variable sets (Table 4). The shared variance on each of the three (1–3) canonical variates are 69.02%, 32.78%, 19.19%, respectively, the canonical variables (CV:s) showed a strong canonical correlation between landcover and land use attributes and water quality for CV1 (0.84), and moderate for CV2 (0.61) and CV3 (0.49). On CV1 structural correlations illustrated strongest correlations between PC1 water quality component with variables indicating low presence of AS soils and high coverage of ditches (Fig. 3A). The variance explained in PC1 was mostly due to these two variables (Fig. 3A). The structural correlations and the variance explained on CV2 showed a connection between PC3 and in the low existence of wetlands in the catchment (Fig. 3B). On the CV3, structural correlations were strongest between PC2 water quality component and higher presence of forested lands and low presence of artificial surfaces areas in the catchment (Fig. 3C). These two catchment variables also had the highest variance explained on PC2. The assemblage structure of fish, benthic invertebrates and diatoms differed among the rivers, and river identity explained a major part of the variation in assemblage structure in all three biological elements (Table 5, Appendix 2). Also, the PCo ordinations indicated clear sepa- ration between the rivers in the assemblage structure for all biotic groups, whereas the assemblage variations among study years were largely overlapping (Fig. 4). This temporal variation, although signifi- cant in fish and diatoms, explained little of the overall assemblage structure of the three biological elements (Table 5). Water quality component PC1 and the area of AS soils correlated with the ordination (assemblage sample spatial medians) of both fish and diatom assemblages on PCo dimension 1 (Table 6). Thus, the fish and diatom assemblage samples that had the least amount of ASS, and water quality least affected by ASS, were located to the left side of the PCo dimension 1 and those with high occurrence and impact to the right (Fig. 4). In addition, fish assemblage structure on PCo1 also correlated with PC2 and diatom assemblage structure with catchment forests cover. On PCo dimension 2, fish assemblages and especially diatom assem- blages, were correlated with PC2 (Table 6). Consequently, to the higher end of the PCo dimension 2 were the sampling sites with water quality characteristics of higher alkalinity, turbidity and Pb content (Fig. 4). To conclude, the PCo ordination of fish and diatom assemblage samples were mostly affected by two water quality components, PC1 and PC2, and coverage of AS soils and forests in the catchment. PCo on benthic invertebrate assemblage structure showed somewhat different main factors structuring the assemblage. The assemblage structure was most strongly correlated with PC2 (PCo dimension 1) and PC3 (dimension 2; Table 6). Thus, benthic invertebrate sampling sites from rivers with low alkalinity and turbidity, and high on color, Fe and Pb were placed on the positive end of horizontal axis (PCo dimension 1), while the ordination the vertical axis seemed to be based on several T. Vehanen et al. Ecological Indicators 141 (2022) 109085 6 Fig. 2. Plot of the Principal Component Analysis (PCA) results on the water quality data (Appendix 1) of the 15 AS soil study rivers. PC components one and two (A) and one and three (B) are shown. Variation explained (%) by each component is given in paren- thesis. The length and the color of the arrows indicate the most important water quality variables contrib- uting to PC components. The squared cosine (cos2) shows the importance of a variable for a given observation: variables with a large value of cos2 contribute a relatively large portion to the PC. T. Vehanen et al. Ecological Indicators 141 (2022) 109085 7 factors. The assemblage structure of benthic invertebrates also corre- lated with coverage of AS soils and forests in the catchment (Table 6). Fish and invertebrate EQR-values were highest in rivers with the highest mean pH (mean river pH pooled over years and study sites, Fig. 5), and both EQRs had lowered along with the decreasing pH values (Spearman rank correlation, fish: rs ˆ 0.525, p ˆ 0.022, N ˆ 15, in- vertebrates: rs ˆ 0.624, p ˆ 0.027, N ˆ 10). Out of the metrics within the invertebrate EQR PMA-index (rs ˆ 0.673, p ˆ 0.017, N ˆ 10) and number of stream type -specific taxa (rs ˆ 0.663, p ˆ 0.018, N ˆ 10) were mostly affected by pH. Within the fish EQR the number of fish species present (rs ˆ 0.757, p < 0.001, N ˆ 15), density of age-0 ‡ salmonids (rs ˆ 0.447, p ˆ 0.048, N ˆ 15) and the density of cyprinid group individuals increased (rs ˆ -0.709, p ˆ 0.002, N ˆ 15, reversed metrics) with rising pH. Diatom EQRs were in general lower than those of fish or invertebrates, (Fig. 5), and the correlation between EQRs and pH values was not significant. The pressure specific diatom index (ACID) (index values, mean ˆ 5.269 range ˆ 3.161–7.609, N ˆ 12) had a sig- nificant correlation with pH (rs ˆ -0.681, p ˆ 0.007, N ˆ 12), indicating relatively strong response by the index to changes in mean pH. All three biological elements, fish, invertebrates, and diatoms had comparable trends in EQR’s against water quality gradient in PC1 that combined the increase in concentrations of several metals to decreasing pH. The EQRs first rapidly decreased but increased in the two rivers with the highest PC1 values (Fig. 5). Because of this the correlation between EQRs and PC1 was significant only in fish (Spearman rank correlation, rs ˆ 0.579, p ˆ 0.012, N ˆ 15). Within the fish index the combination of rising pH with less metals changed the number of fish species present (rs ˆ 0.600, p ˆ 0.009, N ˆ 15), especially increased the proportion of intolerant fish species (rs ˆ 0.623, p ˆ 0.007, N ˆ 15), and also the densities in age-0 ‡ salmonids increased (rs ˆ 0.661, p ˆ 0.004, N ˆ 15). The two rivers breaking the decreasing trend in EQRs (Rivers Laihianjoki and Maalahdenjoki) had the highest metal concentrations (Al, Cd, Co, Cu, Na, Ni, Zn, conductivity), but they were not the most acid rivers in the data (mean pH 5.6 and 4.9). The ACID index values did not correlate significantly with PC1 values (Spearman rank correlation, rs ˆ 0.343, p ˆ 0.138, N ˆ 12), the same two rivers (Rivers Laihianjoki and Maalahdenjoki) breaking the otherwise linear relationship. 4. Discussion We observed a strong relationship between water quality and land- cover attributes, but more variable relations with biological quality el- ements. Several studies have demonstrated how catchment land use interacts with the soil types to impair water quality and often also the aquatic assemblages (Sliva and Williams, 2001; Donohue et al., 2006; Sutela and Vehanen, 2010; Valle Junior et al., 2015; Horak et al., 2020). Although catchment geology and soil types are the major factors determining overall ecology of streams (Hynes, 1975), it is evident that local aquatic assemblage structure is always a result of complex in- teractions between natural and human-induced environmental charac- teristics at multiple spatial and temporal scales (Allan, 2004; Varanka et al., 2015). Our results underline that the key to understand the consequences of this complexity were the responses of the biological communities.. The area of AS soils affected water quality rather straightforwardly and rather much also biological quality. Biological metrics based on fish and invertebrates responded to decreasing pH in rivers with decreasing ecological quality ratios, while diatom metrics showed no significant response to pH. Metrics of all the three biological groups responded negatively to increasing metal concentrations up to moderately impacted rivers, but not in the two rivers with the highest metal con- centrations and not the lowest pH. Thus, it may be that high metal concentrations without very low pH were not especially harmful to the assemblages sampled. Earlier AS river studies have shown that high metal levels (Al, Cd, Zn) may not induce ecotoxicity outside the most adverse acid surges, probably due to the simultaneous high Ca and DOC concentrations lowering metal bioavailability (Vuori 1995b). The level of metals in water quality samples generally surpassed EQS levels indicating lower than good ecological status in AS soils impacted rivers. The interpretation of the effect of metals is not, however, straightfor- ward as their effects are dependent on the bioavailability as a function of several factors (Adams et al., 2020). Our study showed the effect of low pH on metals toxicity and the biological response. The biological quality elements responded differently to the water quality impacts. While the fish and diatom assemblage structures were most strongly affected by the AS soil water quality changes, in- vertebrates were more structured according to the forestry-induced water quality changes, such as water color, turbidity and iron concen- trations. Increased organic loading from the catchment can increase Table 3 Mean yearly levels of metals (µg/l) in the study sites water samples 2010–2012 (N ˆ 126). The Environmental Quality Standards (EQS) values for metal content are given in the last row. The coloring of the mean values refers to: red ˆ numerous times higher than EQS, orange ˆ around twice higher than EQS, yellow ˆ higher than EQS and blue ˆ lower than the EQS criteria. For Na there was no EQS value. *Canadian Freshwater Metals Criteria – CCME, freshwater long term values. **EU Directive 2008/105/EC, annual averages. *** Canadian BC Water Quality Guidelines, freshwater long term values. Table 4 Significance tests (Rao’s F Approximation and Bartlett’s Chi-Squared Test) of the canonical correlation analysis between catchment characteristics and water quality. Canonical variate Canonical correlation Rao’s F statistics Num. df Den. df Prob. (>F) CV1 0.831 13.664 21 333.64 <0.001 CV2 0.572 6.957 12 234.00 <0.001 CV3 0.438 5.605 5 118.00 <0.001 rho^2 Bartlett’s Chisq Df Prob. (>X) CV1 0.690 212.958 21 <0.001 CV2 0.328 72.922 12 <0.001 CV3 0.192 25.465 5 <0.001 T. Vehanen et al. Ecological Indicators 141 (2022) 109085 8 especially the abundance of filter-feeding invertebrates (Karvonen, 1995). The spatial variation in the aquatic assemblages among rivers was clearly higher compared to temporal variation among years. This sug- gests that the factors identified to cause the differences in the assem- blages among the rivers, mainly the diffuse loading from AS soils (pH, metals) and forested areas (color, iron, alkalinity, turbidity), are in the long-term affecting the aquatic assemblages. Thus, to restore the ecological status of these rivers the restoration efforts should be directed to the catchment to diminish these diffuse loads. Previous research has highlighted the importance of mitigation efforts towards the acid leaching from agricultural lands by maintaining the groundwater levels high enough to avoid oxidation of AS soils, but clearly more reduction in acid loadings is needed to reach the target of good ecological status (Vintanen et al., 2016) Restoration of the disturbed freshwater ecosys- tems requires deactivating and mitigating the same mechanisms that have caused the ecosystem damage (Rapport and Whitford, 1999). Our results demonstrate that the large area of AS soils had pro- nounced impacts on acidity and metal concentrations in rivers. Diffuse loading from agriculture on AS soils is the main source of leaching of metals to waters in Finland (Åstrom and Bjorklund, 1995). Often species diversity tends to decrease, and composition change in areas with elevated metal concentrations, but the mechanisms are not well known (Mayer-Pinto et al., 2010). According to Wallin et al. (2015) the con- centration of metals in the AS soil river estuaries can be so high that harmful ecotoxicological impacts are likely to occur. For fish pop- ulations, the negative effect of environmental acidification is especially linked to aluminum (Al), as the combined effect of low pH and high Al level damages gill tissue (Youson and Neville, 1987; Muniz and Lei- vestad, 1980; Exley et al., 1991). Indeed, high Al levels combined with acidity in Finnish AS soil rivers were evaluated to affect fish assemblage structure and diminish species richness (Sutela & Vehanen 2017). In our study, elevated concentrations were not limited to Al, but also Cd, Co, Ni and Zn concentrations increased with increased proportion of AS soils and low pH. Overall, the combined impact of metals to aquatic assem- blages remains poorly known. Given that more AS soils are increasingly taken to anthropogenic use (Smith et al., 2016), the metal pollution of surface and groundwaters is expected to accelerate in the future (Dent Fig. 3. The Canonical Correlation Analysis results of the associations between water quality (the three first water quality principal components, see Fig. 2) to landcover attributes in 15 AS soil rivers. The uppermost panel (A) shows the first Canonical Variable (CV), mid panel (B) is the second CV and the lowermost panel (C) is the third CV. The right panel shows the explained variance for by each variable (length of the black bar indicates the amount variance explained) and the left panel the structural correlations (black bars are positive correlations, white bars negative) for each variable for the CV in question. T. Vehanen et al. Ecological Indicators 141 (2022) 109085 9 and Pons 1995; Hinwood et al., 2008; Enio et al., 2020); hence, effective biological assessment tools are needed. The impacts of metals are not always directly related to acidification. In our study this was the case with Fe, Pb, As and Cr. Similar to our results, Åstrom and Bjorklund (1995) reported that concentrations of some metals like Al, Co, and Ni increase due to runoff from AS soils, but others, like Fe, do not increase above their baseline concentrations due to AS soils. The reason for this can be in oxidation–reduction processes, Fe, for example, can be rapidly immobilized by oxidation processes (Åstrom and Bjorklund, 1995). One plausible explanation is that these metals interact differently with humic compounds when water becomes more acidic. Vuori (1995a) observed a dramatic decrease of Fe con- centrations during the acid surges, which was related to the co- precipitation of iron with humic compounds. These precipitates may increase mortality of invertebrates and fish embryos and induce indirect ecosystem changes (Vuori, 1995a). High concentrations of iron may also cause direct mortality: Myllynen et al. (1997) reported lower survival of river lamprey (Lampetra fluviatilis) larvae when total iron concentrations rose to high levels representing natural high flow conditions of an AS soil river. Fe has a prevalent role in ecosystem functioning in boreal fresh- waters, and there is an indication that Fe concentrations have recently increased in boreal freshwaters, probably due to changes in precipita- tion and temperature (Heikkinen et al., 2022). The most dramatic effects of agricultural AS soils emerge when heavy rainfall periods and elevated runoff are followed by prolonged dry periods with lowered groundwater levels enabling oxidation of AS soils and effective release of acidity (Toivonen and Osterholm, 2011; Nystrand and Osterholm, 2013). Such acid peaks have caused fish kills in several rivers in our study area (Sutela et al., 2012). During our study period there were no such sudden acidic peaks. Instead, rivers main- tained their typical pH levels related especially to the quantity of AS soils, and degree of human disturbance. It is clear that acidity and metal Table 5 Summary of PERMANOVA statistics on Bray-Curtis dissimilarities of fish, diatom and benthic invertebrate assemblages among study rivers, study years and their interaction. The results are based on 999 permutations in each case. Fish Df Sum. of sgs. Mean sgs. Pseudo- F r2 Prob. (>F) River 14 21.264 1.519 10.925 0.565 0.001 Year 1 0.370 0.370 2.661 0.010 0.027 River*Year 14 2.638 0.188 1.355 0.070 0.042 Residuals 96 15.985 0.139 0.423 Total 125 37.619 1.000 Diatoms Df Sum. of sgs. Mean sgs. Pseudo- F r2 Prob. (>F) River 12 15.419 1.285 8.594 0.522 0.001 Year 1 0.686 0.686 4.588 0.023 0.001 River*Year 12 2.070 0.172 1.154 0.070 0.138 Residuals 76 11.364 0.149 0.385 Total 101 29.539 1.000 Invertebrates Df Sum. of sgs. Mean sgs. Pseudo- F r2 Prob. (>F) River 9 4.178 0.464 1.853 0.619 0.007 Year 1 0.169 0.169 0.675 0.025 0.756 River*Year 9 1.153 0.128 0.512 0.171 1.000 Residuals 5 1.253 0.250 0.185 Total 24 6.753 1.000 Fig. 4. Distances (Bray Curtis dissimilarities) between study rivers (left panel) and years (right panel) on Principal Coordinates Analysis (PCoA) dimensions one and two. The uppermost panel shows the results based on fish assemblage, mid panel based on diatom assemblage and the lowermost panel is based on invertebrate assemblage. River name or year shows the spatial median and the line covers the area of the samples for each group. T. Vehanen et al. Ecological Indicators 141 (2022) 109085 10 load strongly vary spatially and temporally, depending on the amount of AS soils and their degree of oxidation which can be largely affected by land use, especially cultivation and drainage intensity (Vintanen et al., 2016). This variation was apparent also in our ecological quality assessment results. Our results showed that the main drivers behind the ecosystem degradation in the study rivers were acidity pressures related to AS soils in the catchment. The national metrics used in this study where not acidity-specific, but were originally developed to respond to impacts of most common pressures, especially eutrophication and hydro- morphological alteration (Vehanen et al., 2010; Aroviita et al., 2008; Aroviita et al., 2012). In our study fish and invertebrate metrics responded to pH gradient, while diatom metrics did not. To effectively target the mitigation measures it is important that biological indices respond to the key pressures of the area (Rapport and Hilden, 2013). If not, the obvious consequence should be to develop indices that are more specific to specific pressure impacts, as it is the case in the diatoms for which ACID index was developed by Andren and Jarlman (2008). Ac- cording to our results the ACID index responded well to the changes in pH in AS soils rivers. In the case of the effect of increasing metal con- centrations the results were rather unified among the three biological metrics: the EQR values first steeply decreased as expected, but then slightly increased along the PC1 axis with metal concentration increase but also water pH decrease. It remains unclear whether in this case at the end of PC1 axis the signal of better ecological status was true, or falsely positive in failing to indicate potential stress from the high metal content. The obvious reason for the diatom metrics not to respond to the ef- fects of low pH is that the metrics (type-specific taxa and PMA) were applied for Finnish rivers to reflect mainly impacts of pressures such as Table 6 Spearman rank correlation coefficients (rs) between assemblage structure of fish, benthic invertebrates and diatoms and Principal Coordinates Analysis (PCoA) results with catchment and land use characteristics of the rivers studied. The spatial medians of the fish, diatom and benthic invertebrate samples were used for PCo dimension one and two. The most significant correlations (p < 0.001) are indicated by bold typeface. Fish PC1 PC2 PC3 Forests Wetlands Waters AS soils Ditches Agriculture Artificial surfaces PCo dimension1 ¡0.666 0.400 0.254 0.278 0.245 0.101 0.512 0.303 0.058 0.071 Prob. <0.001 <0.001 0.004 0.002 0.005 0.259 <0.001 0.001 0.517 0.428 PCo dimension2 0.034 ¡0.398 0.242 0.356 0.174 0.218 0.105 0.214 0.225 ¡0.350 Prob. 0.703 <0.001 0.006 <0.001 0.051 0.014 0.243 0.016 0.011 <0.001 Diatoms PCo dimension1 ¡0.504 0.118 0.215 0.466 0.207 0.114 0.373 0.221 0.156 0.230 Prob. <0.001 0.236 0.030 <0.001 0.037 0.255 <0.001 0.026 0.118 0.020 PCo dimension2 0.160 ¡0.667 0.061 0.044 0.003 0.144 0.050 0.151 0.031 0.174 Prob. 0.108 <0.001 0.546 0.664 0.977 0.255 0.619 0.130 0.759 0.081 Invertebrates PCo dimension1 0.196 0.628 0.134 0.194 0.067 0.260 0.190 0.250 0.022 0.201 Prob. 0.479 0.001 0.522 0.347 0.751 0.209 0.363 0.228 0.918 0.334 PCo dimension2 0.149 0.299 ¡0.466 ¡0.477 0.050 0.424 0.445 0.405 0.321 0.413 Prob. 0.479 0.149 0.019 0.016 0.814 0.035 0.026 0.045 0.118 0.040 Fig. 5. Relationship between mean pH (upper panel) and water quality gradient (Principal Component one, lower pamel) to Ecological Quality Ratio calculated from fish, invertebrate and diatom data of the study rivers. Data is pooled over rivers and study years (2010–2012, mean  SD). T. Vehanen et al. Ecological Indicators 141 (2022) 109085 11 nutrient leaching and organic pollution. Mykra et al. (2019) reported that in streams with natural acidic environment nutrient enrichment did not increase diatom richness as it did in neutral environment. Hence the acidity may override eutrophication effects on diatom community. Ac- cording to our results impacts from the AS soils altered the composition of invertebrate taxa compared to rivers with less acid pressures. Increased organic loading from the catchment can increase especially the abundance of filter-feeding invertebrates (Karvonen, 1995), which partly explain the observed partly irregular assemblage and metric patterns along water quality and pH gradient. Earlier studies have shown that invertebrates show species-specific differences in sensitivity to AS soil impacts (Vuori 1995b; Willner 2020). Runoff from AS soils has caused slower larval development and increased the occurrence of ab- normalities among invertebrates (Vuori 1995b). Our fish results showed that acidity decreased the number of species present and ceased the reproduction of salmonids. The combination of decreasing pH and increasing metal content affected the fish community structure by decreasing the proportion of intolerant fish species. Somewhat similarly, many of the key fish species in Finnish rivers classified as intolerant in FiFI index (e.g., brown trout, grayling Thymallus thymallus and European bullhead Cottus gobio) do not tolerate acidity (Sutela and Vehanen, 2017) suggesting that FiFI index could be applicable with rivers disturbed by acidification. However, cyprinids are classified as tolerant species in FiFI though their tolerance to acidification is poor (Almer et al., 1974; Rahel and Magnuson, 1983). This was also confirmed by our results, as their density decreased with decreasing pH. In this aspect, the sensitivity of the metrics towards detecting acidification pressures could be further developed. Overall, our results support the conclusions of earlier studies on the necessity and added value to include several bio- logical assemblages in the assessment of ecological impacts in rivers also when it comes to the effects from acid releases (Bae et al., 2011, 2014; Pace et al., 2012; de Morais et al., 2018). Declaration of Competing Interest The authors declare that they have no known competing financial interest or personal relationship that could have appeared to influence the work reported in this paper. Data availability Data will be made available on request. Acknowledgements We wish to thank all people that helped us in the field work. This work was supported by the CATERMASS (Climate Change Adaptation Tools for Environmental Risk Mitigation of Acid Sulphate Soils) Life ‡ -project (LIFE08 ENV/FI/000609),), the Nordic Centre of Excellence “BIOWATER” (Nordforsk Project no. 82263) and by the Ministry of Agriculture and Forestry as part of the monitoring network MaaMet. Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi. org/10.1016/j.ecolind.2022.109085. References Abdi, H., Williams, L.J., 2010. Principal Component Analysis. Wiley Interdiscip. Rev. Comput. Stat. 2, 433–459. https://doi.org/10.1002/wics.101. Adams, W., Blust, R., Dwyer, R., Mount, D., Nordheim, E., Rodriguez, P.H., Spry, D., 2020. Bioavailability assessment of metals in freshwater environments: A historical review. Environmental Toxicology and Chemistry 39, 48–59. https://doi.org/ 10.1002/etc.455. Allan, J.D., 2004. Landscapes and Riverscapes: The Influence of Land Use on Stream Ecosystems. Annu. Rev. Ecol. Evol. Syst. 35, 257–284. https://doi.org/10.1146/ annurev.ecolsys.35.120202.110122. Allen, A.P., Whittier, T.R., Larsen, D.P., Kaufmann, P.R., O’Connor, R.J., Hughes, R.M., Stemberger, R.S., Dixit, S.S., Brinkhurst, R.O., Herlihy, A.T., Paulsen, S.G., 1999. Less Concordance of taxonomic composition patterns across multiple lake assemblages: effects of scale, body size, and land use. Can. J. Fish. Aquat. Sci. 56 (11), 2029–2040. Almer, B., Dickson, W., Ekstrom, C., Hornstrom, E., 1974. Effects of acidification on Swedish lakes. Ambio 3, 30–36. Anderson, M.J., 2001. A new method for non-parametric multivariate analysis of variance. Austral Ecol. 26, 32–46. https://doi.org/10.1111/j.1442- 9993.2001.01070. pp. x. Andren, C., Jarlman, A., 2008. Benthic diatoms as indicators of acidity in streams. Fundam. Appl. Limnol. Arch. Hydrobiol. 173 (3), 237–253. Andriesse, W., van Mensvoort, M.E.F., 2002. Distribution and extent of acid sulphate soils. In: Lal, R. (Ed.), Encyclopedia of soil science. Marcel Dekker Inc., New York, pp. 1–6. Aroviita, J., Hellsten, S., Jyvasjarvi, J., Jarvenpaa, L., Jarvinen, M., Karjalainen, S.M., Kauppila, P., Keto, A., Kuoppala, M., Manni, K., Mannio, J., Mitikka, S., Olin, M., Pilke, A., Rask, M., Juha, R., Sutela, T., Vehanen, T., Vuori, K.-M., 2012. Ohje pintavesien ekologisen ja kemiallisen tilan luokitteluun vuosille 2012–2013 — paivitetyt arviointiperusteet ja niiden soveltaminen. (Ymparistohallinnon ohjeita). Suomen ymparistokeskus 7. http://www.ymparisto.fi/default.asp? contentidˆ427506&lanˆfi. Aroviita, J., Mitikka, S., Vienonen S., (ed.) 2019. Status classification and assessment criteria of surface waters in the third river basin management cycle. Suomen ymparistokeskuksen raportteja 37/2019, 1–177 (in Finnish). http://hdl.handle.net/ 10138/306745. Aroviita, J., Koskenniemi, E., Kotanen, J., Hamalainen, H., 2008. A priori typology-based prediction of benthic macroinvertebrate fauna for ecological classification of rivers. Environ. Manag. 42 (5), 894–906. Åstrom, M., Bjorklund, A., 1995. Impact of acid sulfate soils on stream water geochemistry in western Finland. J. Geochem. Explor. 55 (1-3), 163–170. Åstrom, M., Bjorklund, A., 1996. Hydrogeochemistry of a stream draining sulfde-bearing postglacial sediments in Finland. Water, Air Soil Pollut. 89, 233–246. Bae, M.J., Li, F., Kwon, Y.S., Chung, N., Choi, H., Hwang, S.J., Park, Y.S., 2014. Concordance of diatom, macroinvertebrate and fish assemblages in streams at nested spatial scales: Implications for ecological integrity. Ecol. Indic. 47, 89–101. Bae, M.J., Kwon, Y., Hwang, S.J., Chon, T.S., Yang, H.J., Kwak, I.S., Park, J.H., Ham, S. A., Park, Y.S., 2011. Relationships between three major stream assemblages and their environmental factors in multiple spatial scales. Ann. Limnol. 47, 91–105. Bayramoglu, B., Chakir, R., Lungarska, A., 2020. Impacts of Land Use and Climate Change on Freshwater Ecosystems in France. Environ. Model. Assess. 25 (2), 147–172. Boman, A., Frojdo, S., Backlund, K., Åstrom, M.E., 2010. Impact of isostatic land uplift and artificial drainage on oxidation of brackish-water sediments rich in metastable iron sulfide. Geochim. Cosmochim. Acta 74 (4), 1268–1281. Butts, C.T., 2018. Yacca: Yet Another Canonical Correlation Analysis Package. R package version 1 (1), 1. Carlisle, D.M., Hawkins, C.P., Meador, M.R., Potapova, M., Falcone, J.A., 2008. Biological assessments of Appalachian streams based on predictive models for fish, macroinvertebrate, and diatom assemblages. J. North Am. Benthol. Soc. 27, 16–37. https://doi.org/:10.1899/06-081.1. Chen, Y., Qu, X., Xiong, F., Lu, Y., Wang, L., Hughes, R.M., 2020. Challenges to saving China’s freshwater biodiversity: Fishery exploitation and landscape pressures. Ambio 49 (4), 926–938. Cook, F.J., Hicks, W., Gardner, E.A., Carlin, G.D., Froggatt, D.W., 2000. Export of acidity in drainage water from Acid Sulphate Soils. Mar. Pollut. Bull. 41 (7-12), 319–326. Corfield, J., 2000. The effects of acid sulphate run-offon a subtidal estuarine macrobenthic community in the Richmond River, NSW. Australia. ICES J. Mar. Sci. 57 (5), 1517–1523. Cross, W.F., Baxter, C.V., Rosi-Marshall, E.J., Hall, R.O., Kennedy, T.A., Donner, K.C., Wellard Kelly, H.A., Seegert, S.E.Z., Behn, K.E., Yard, M.D., 2013. Food-web dynamics in a large river discontinuum. Ecol. Monogr. 83 (3), 311–337. Dent, D.L., Pons, L.J., 1995. A world perspective on acid sulphate soils. Geoderma 67 (3- 4), 263–276. Donohue, I., McGarrigle, M.L., Mills, P., 2006. Linking catchment characteristics and water chemistry with the ecological status of Irish rivers. Water Res. 40 (1), 91–98. Dudgeon, D., Arthington, A.H., Gessner, M.O., Kawabata, Z.-I., Knowler, D.J., Leve^que, C., Naiman, R.J., Prieur-Richard, A.-H., Soto, D., Stiassny, M.L.J., Sullivan, C.A., 2006. Freshwater biodiversity: importance, threats, status and conservation challenges. Biol. Rev. 81, 163–182. Enio, M.S.K., Shamshuddin, J., Fauziah, C.I., Husni, M.H.A., Panhwar, Q.A., 2020. Quantifying the release of acidity and metals arising from drainage of acid sulfate soils in the Kelantan Plains. Malaysia. Min. Miner. Depos. 14 (3), 50–60. EU Directive 2008/105/EC of the European Parliament and of the Council of 16 December 2008 on environmental quality standards in the field of water policy, amending and subsequently repealing Council Directives 82/176/EEC, 83/513/EEC, 84/156/EEC, 84/491/EEC, 86/280/EEC and amending Directive 2000/60/EC of the European Parliament and of the Council. Official Journal of the European Union L 348/84. Exley, C., Chappell, J.S., Birchall, J.D., 1991. A mechanism for acute aluminium toxicity in fish. J. Theor. Biol. 151 (3), 417–428. Feio, M.J., Almeida, S.F.P., Aguiar, F.C., 2017. Functional associations between microalgae, macrophytes and invertebrates distinguish river types. Aquat. Sci. 79 (4), 909–923. T. Vehanen et al. Ecological Indicators 141 (2022) 109085 12 Faltmarsch, R.M., Åstrom, M.E., Vuori, K.-M., 2008. Environmental risks of metals mobilized from acid sulphate soils in Finland: a literature review. Boreal Environ. Res. 13, 444–456. Grenouillet, G., Brosse, S., Tudesque, L., Lek, S., Baraille, Y., Loot, G., 2008. Concordance among stream assemblages and spatial autocorrelation along a fragmented gradient. Divers. Distrib. 14, 592–603. Gupta, V.V.S.R., Germida, J.J., 2021. Microbial transformations of sulfur in soil. In: Gentry, T.J., Fuhurmann, J.J., Zuberer, D.A. (Eds.), Principles and Applications of Soil Microbiology. Elsevier, Amsterdam, pp. 493–522. Hautier, Y., Tilman, D., Isbell, F., Seabloom, E.W., Borer, E.T., Reich, P.B., 2015. Anthropogenic environmental changes affect ecosystem stability via biodiversity. Science 348 (6232), 336–340. Heikkinen, K., Saari, M., Heino, J., Ronkanen, A.-K., Kortelainen, P., Joensuu, S., Vilmi, A., Karjalainen, S.-M., Hellsten, S., Visuri, M., Marttila, H., 2022. Iron in boreal river catchments: Biogeochemical, ecological and management implications. Sci. Total Environ. 805, 150256. Hinwood, A., Horwitz, P., Rogan, R., 2008. Human exposure to metals in groundwater affected by acid sulfate soil disturbance. Arch. Environ. Contam. Tocicol. 55 (3), 538–545. Horak, C.N., Assef, Y.A., Grech, M.G., Miserendino, M.L., 2020. Agricultural practices alter function and structure of macroinvertebrate communities in Patagonian piedmont streams. Hydrobiologia 847 (17), 3659–3676. Horn, J.L., 1965. A rationale and test for the number of factors in factor analysis. Psychometrica 30 (2), 179–185. Hynes, H.B.N., 1975. The Stream and its Valley. Verh. - Int. Ver. Theor. Angew. Limnol. 19 (1), 1–15. Jackson, D.A., Harvey, H.H., 1993. Fish and benthic invertebrates: Community concordance and community-environment relationships. Can. J. Fish. Aquat. Sci. 50, 2641–2651. https://https://doi.org/.org/10.1139/f93-287. Juggins, S., Kelly, M., Allott, T., Kelly-Quinn, M., Monteith, D.T., 2016. A Water Framework Directive-compatible metric for assessing acidification in UK and Irish rivers using diatoms. Sci. Total Environ. 568, 671–678. https://doi.org/10.1016/j. scitotenv.2016.02.163. Jarvinen, M., Aroviita, J., Hellsten, S., Karjalainen, S.M., Kuoppala, M., Meissner, K., Mykra, H., Vuori, K.-M., 2020. Jokien ja jarvien biologinen seuranta – naytteenotosta tiedon tallentamiseen. Ver 6.9.2019. 42 pp. Suomen ymparistokeskus. https://www.ymparisto.fi/fiFI/Vesi/Pintavesien_tila/Pintavesien_ tilan_-seuranta/Biologisten_seurantamentelmien_ohjeet. Karvonen, K., 1995. The Effects of Peat Mining on Fluvial Benthic Fauna. Licentiate’s theses. University of Oulu, Department of. Zoology. 71, pp in Finnish. Kassambara, A., Mundt, F., 2020. Factoextra: Extract and Visualize the Results of Multivariate Data Analyses. R Package Version 1.0.7. https://CRAN.R-project.org/ packageˆfactoextra. Kilgour, B.W., Barton, D.R., 1999. Associations between stream fish and benthos across environmental gradients in southern Ontario. Canada. Freshw. Biol. 41, 553–566. Mayer-Pinto, M., Underwood, A.J., Tolhurst, T., Coleman, R.A., 2010. Effects of metals on aquatic assemblages: What do we really know? J. Exp. Mar. Biol. Ecol. 391 (1-2), 1–9. de Morais, G.F., Ribas, L.G.D., Ortega, J.C.G., Heino, J., Bini, L.M., 2018. Biological surrogates: A word of caution. Ecol. Indic. 88, 214–218. Muniz, I.P., Leivestad, H., 1980. Acidifcation effects on freshwater fsh. In: Drabløs, D., Tollan, A. (Eds.), Ecological impact of acid precipitation. SNSF project, Oslo-Ås, Norway, pp. 84–92. Myllynen, K., Ojutkangas, E., Nikinmaa, M., 1997. River water with high iron concentration and low pH causes mortality of lamprey roe and newly hatched larvae. Ecotoxicol. Environ. Saf. 36 (1), 43–48. Mykra, H., Sarremejane, R., Laamanen, T., Karjalainen, S.M., Markkola, A., Lehtinen, S., Lehosmaa, K., Muotka, T., 2019. Local geology determines responses of stream producers and fungal decomposers to nutrient enrichment: A field experiment. Ambio 48, 100–110. https://doi.org/10.1007/s13280-018-1057-4. Novak, M.A., Bode, R.W., 1992. Percent model affinity: A new measure of macroinvertebrate community composition. J. North Am. Benthol. Soc. 11 (1), 80–85. Nystrand, M.I., Osterholm, P., 2013. Metal species in a boreal river system affected by acid sulfate soils. Appl. Geochem. 31, 133–141. Ministry of Agriculture and Forestry and Ministry of the Environment, 2011. Guidelines for mitigating the adverse effects of acid sulphate soils in Finland until 2020. Publications of Ministry of Agriculture and Forestry 2/2011. 26 p. Oksanen, J., Blanchet, F.G., Friendly, M., Kindt, R., Legendre, P., McGlinn, D., Minchin, P.R., O’Hara, R.B., Simpson, G.L., Solymos, P., Stevens, M.H.H., Szoecs, E., Wagner, H., 2019. vegan: Community Ecology Package. R package version 2.5-6. http:// CRAN.Rproject.org/packageˆvegan. Pace, G., Della Bella, V., Barile, M., Andreani, P., Mancini, L., Belfiore, C., 2012. A comparison of macroinvertebrate and diatom responses to anthropogenic stress in small sized volcanic siliceous streams of Central Italy (Mediterranean Ecoregion). Ecol. Indic. 23, 544–554. Poikane, S., Salas Herrero, F., Kelly, M.G., Borja, A., Birk, S., van de Bund, W., 2020. European aquatic ecological assessment methods: A critical review of their sensitivity to key pressures. Sci.Total Environ. 740, 140075. Poikane, S., Zampoukas, N., Borja, A., van de Davies, S.P., Bund, W., Birk, S., 2014. Intercalibration of aquatic ecological assessment methods in the European Union: Lessons learned and way forward. Environ. Sci. Policy 44, 237–246. Pont, D. (coordinator), 2011. Water Framework Directive. Intercalibration Phase 2. River Fish European Intercalibration Group. Final Report to ECOSTAT. Popp, A., Calvin, K., Fujimori, S., Havlik, P., Humpenoder, F., Stehfest, E., Bodirsky, B.L., Dietrich, J.P., Doelmann, J.C., Gusti, M., Hasegawa, T., Kyle, P., Obersteinerd, M., Tabeau, A., Takahashi, K., Valin, H., Waldhoff, S., Weindl, I., Wise, M., Kriegler, E., Lotze-Campen, H., Fricko, O., Riahi, K., van Vuuren, D.P., 2017. Land-use futures in the shared socio-economic pathways. Glob. Environ. Change 42, 331–345. R Core Team, 2019. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria https://www.R-project.org/. Rahel, F.J., Magnuson, J.J., 1983. Low pH and the absence of fish species in naturally acidic Wisconsin lakes: Inferences for cultural acidification. Can. J. Fish. Aquat. Sci. 40 (1), 3–9. Rapport, D.J., Hilden, M., 2013. An evolving role for ecological indicators: From documenting ecological conditions to monitoring drivers and policy responses. Ecol. Indic. 28, 10–15. Rapport, D.J., Whitford, W.G., 1999. How Ecosystems Respond to Stress: Common Properties of Arid and Aquatic Systems. BioScience 49, 193–203. Reid, A.J., Carlson, A.K., Creed, I.F., Eliason, E.J., Gell, P.A., Johnson, P.T.J., Kidd, K.A., MacCormack, T.J., Olden, J.D., Ormerod, S.J., Smol, J.P., Taylor, W.W., Tockner, K., Vermaire, J.C., Dudgeon, D., Cooke, S.J., 2019. Emerging threats and persistent conservation challenges for freshwater biodiversity. Biol. Rev. 94 (3), 849–873. Rosi-Marshall, E.J., Wallace, J.B., 2002. Invertebrate food webs along a stream resource gradient. Freshw. Biol. 47, 129–141. Ritsema, C.J., van Mensvoort, M.E.F., Dent, D.L., Tan, Y., van den Bosch, H., van Vijk, A. L.M., 2000. Acid sulfate soils, in: Sumner M.E. (Ed.), Handbook of soil science, CRC press, Boca Raton, pp. 212–154. Smith, P., House, J.I., Bustamante, M., Sobocka, J., Harper, R., Pan, G., West, P.C., Clark, J.M., Adhya, T., Rumpel, C., Paustian, K., Kuikman, P., Cotrufo, F.M., Elliott, J.A., McDowel, R., Griffiths, R.I., Asakawa, S., Bondeau, A., Jain, A.K., Meersmans, J., Pugh, T.A.M., 2016. Global change pressures on soils from land use and management. Glob. Chang. Biol. 2016 (22), 1008–1028. Soininen, J., Paavola, R., Kwandrans, J., Muotka, T., 2009. Diatoms: unicellular surrogates for macroalgal community structure in streams? Biodivers. Conserv. 18 (1), 79–89. Sutela, T., Vehanen, T., 2010. Responses of fluvial fish assemblages to agriculture within the boreal zone. Fish. Manag. Ecol. 17, 141–145. Sutela, T., Vehanen, T., 2017. The effects of acidity and aluminium leached from acid- sulphate soils on riverine fish assemblages. Boreal Env. Res. 22, 385–391. Sutela, T., Vehanen, T., Jounela, P., 2020. Longitudinal patterns of fish assemblages in European boreal streams. Hydrobiologia 847 (15), 3277–3290. Sutela, T., Vuori, K.-M., Louhi, P., Hovila, K., Jokela, S., Karjalainen, S.-M., Keinanen, M., Rask, M., Teppo, A., Urho, L., Vehanen, T., Vuorinen, P.J., Osterholm, P., 2012. The impact of acid sulphate soils on water bodies and fish deaths in Finland. Suomen Ymparisto 14/2012. 61 pp. (In Finnish with English summary) https://doi.org/ 10.13140/RG.2.2.29227.90401. Tikkanen, M., Oksanen, J., 2002. Late Weichselian and Holocene shore displacement history of the Baltic Sea in Finland. Fennia 180, 9–20. Toivonen, J., Hudd, R., Nystrand, M., Osterholm, P., 2020. Climatic effects on water quality in areas with acid sulfate soils with commensurable consequences on the reproduction of burbot (Lota lota L.). Environ. Geochem. Health. 42 (10), 3141–3156. Toivonen, J., Osterholm, P., 2011. Characterization of acid sulfate soils and assessing their impact on a humic boreal lake. J. Geochem. Explor. 110 (2), 107–117. Valle Junior, R.F., Varandas, S.G.P., Pacheco, F.A.L., Pereira, V.R., Santos, C.F., Cortes, R. M.V., Sanches Fernandes, L.F., 2015. Impacts of land use conflicts on riverine ecosystems. Land Use Policy 43, 48–62. Varanka, S., Hjort, J., Luoto, M., 2015. Geomorphological factors predict water quality in boreal rivers. Earth Surf. Process. Landforms 40 (15), 1989–1999. Vehanen, T., Sutela, T., Harjunpaa, A., 2020. The effects of ecoregions and local environmental characteristics on spatial patterns in boreal riverine fish assemblages. Ecol. Freshw. Fish 29 (4), 739–751. Vehanen, T., Sutela, T., Korhonen, H., 2010. Environmental assessment of boreal rivers using fish data – a contribution to the Water Framework Directive. Fish. Manag. Ecol. 17, 165–175. Vintanen, S., Uusi-Kamppa, J., Osterholm, P., Bonde, A., Yli-Halla, M., 2016. Potential of controlled drainage and sub-irrigation to manipulate groundwater table for mitigating acid loadings in Finnish acid sulfate soils. Conference Proceedings of the 10th International Drainage Symposium 2016: 372-378. Minneapolis, United States, 6.9-9.9.2016. American Society of Agricultural and Biological Engineers. https:// doi.org/:10.13031/ids.20162521555. Vuori, K.-M., 1995a. Direct and indirect effects of iron on river ecosystems. Ann. Zool. Fennici. 32, 317–321. Vuori, K.-M., 1995b. Species- and population-specific responses of translocated hydropsychid larvae (Trichoptera, Hydropsychidae) to runoff from acid sulphate soils in the river Kyronjoki, Western Finland. Freshw. Biol. 33, 305–318. Wallin, J., Karjalainen, A.K., Schultz, E., Jarvisto, J., Leppanen, M., Vuori, K.-M., 2015. Weight-of-evidence approach in assessment of ecotoxicological risks of acid sulphate soils in the Baltic Sea river estuaries. Sci. Total Environ. 508, 452–461. Weenink, D., 2003. Canonical correlation analysis. Institute of Phonetic Sciences, University of Amsterdam, Proceedings 25, 81–99. Willner, M., 2020. Macroinvertebrates as bioindicators in acid sulfate soil affected streams : Case study of river Perhonjoki catchment area. Åbo Academi University. Masters Thesis. Woodward, G., Perkins, D.M., Brown, L.E., 2010. Climate change and freshwater ecosystems: Impacts across multiple levels of organization. Philos. Trans. R. Soc. Lond., B. Biol. Sci. 365 (1549), 2093–2106. T. Vehanen et al. Ecological Indicators 141 (2022) 109085 13 Yli-Halla, M., Puustinen, M., Koskiaho, J., 1999. Area of cultivated acid sulfate soils in Finland. Soil Use Manag. 15, 62–67. https://doi.org/:10.1111/j.1475-2743.1999. tb00065.x. Youson, J.H., Neville, C.M., 1987. Deposition of aluminium in the gill epithelium of rainbow trout, (Salmo gairdneri) Richardson subjected to sublethal concentrations of the metal. Can. J. Zool. 65, 647–656. Osterholm, P., Åstrom, M., 2002. Spatial trends and losses of major and trace elements in agricultural acid sulphate soils distributed in the artificially drained Rintala area. W. Finland. Appl. Geochem. 17 (9), 1209–1218. T. Vehanen et al.