Published November 10, 2014

Journal of Environmental Quality

TECHNICAL REPORTS Landscape and Watershed Processes

The Use of Phytometers for Evaluating Restoration Effects on Riparian Soil Fertility Anna L. Dietrich, Lovisa Lind, Christer Nilsson,* and Roland Jansson

R

iparian zones, or the areas along streams or rivers

Abstract

that are impacted by flooding and high water tables, link terrestrial and aquatic ecosystems both longitudinally and laterally and through hyporheic exchange (Ward, 1989; Boulton et al., 1998; Baxter et al., 2005). Riparian zones provide important ecosystem functions as they support high levels of biodiversity, buffer aquatic temperature variation, and supply organic matter to aquatic food webs (Naiman et al., 1993; Nilsson et al., 1994; Helfield et al., 2007). Riparian zones can also be viewed as hotspots for biogeochemical processes, such as denitrification, which reduces nitrogen loads before reaching aquatic ecosystems (Hill, 1996; McClain et al., 2003; Gift et al., 2010; Unghire et al., 2011). Worldwide, riparian zones are among the most impacted and altered ecosystems because of factors such as fragmentation, flow regulation, and habitat modification (Maddock, 1999; Malmqvist and Rundle, 2002; Nilsson et al., 2005b; Dudgeon et al., 2006). For example, many streams and rivers that were used for timber floating in the 19th and 20th centuries in Sweden and other northern regions were simplified in terms of channel morphology and flow regime, leading to decreased hydrological connectivity between the riparian and in-stream areas (Törnlund and Östlund, 2002; Nilsson et al., 2005a; Muotka and Syrjänen, 2007). Such alterations can fundamentally alter ecosystem processes and functions, resulting in reduced habitat availability, more erosion and less deposition, and lower retention capacity of litter and propagules (Naiman et al., 1998; Nilsson et al., 2005a; Helfield et al., 2007). To counteract such alterations, restoration, or “the process of assisting the recovery” of a degraded ecosystem (SER, 2004), has become increasingly important worldwide, not least in stream and wetland ecosystems (Bernhardt et al., 2005; Palmer et al., 2005; Roni et al., 2008). In channelized streams, restoration measures often aim at returning boulders and cobbles along with in-stream wood to the channel to slow down current velocity, thus increasing channel width and water-level fluctuations, as well as the width and hydrological connectivity of the riparian zone (Unghire et al., 2011; Gardeström et al., 2013; Nilsson et al., 2014). Such restoration efforts are expected to change riparian plant community composition directly by altering the disturbance regime mediated by floods and ice, as well as indirectly by enhancing sediment and organic matter circulation

The ecological restoration of streams in Sweden has become increasingly important to counteract effects of past timber floating. In this study, we focused on the effect on riparian soil properties after returning coarse sediment (cobbles and boulders) to the channel and reconnecting riparian with instream habitats. Restoration increases habitat availability for riparian plants, but its effects on soil quality are unknown. We also analyzed whether the restoration effect differs with variation in climate and stream size. We used standardized plant species to measure the performance of a grass (Phleum pratense L.) and a forb (Centaurea cyanus L.) in soils sampled in the riparian zones of channelized and restored streams and rivers. Furthermore, we analyzed the mass fractions of carbon (C) and nitrogen (N) along with the proportions of the stable isotopes 13C and 15N in the soil, as well as its grain size composition. We found a positive effect of restoration on biomass of phytometers grown in riparian soils from small streams, indicating that restoration enhanced the soil properties favoring plant performance. We suggest that changed flooding with more frequent but less severe floods and slower flows, enhancing retention, could explain the observed patterns. This positive effect suggests that it may be advantageous to initiate restoration efforts in small streams, which make up the highest proportion of the stream network in a catchment. Restoration responses in headwater streams may then be transmitted downstream to facilitate recovery of restored larger rivers. If the larger rivers were restored first, a slower reaction would be expected.

Copyright © American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America. 5585 Guilford Rd., Madison, WI 53711 USA. All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. J. Environ. Qual. 43:1916–1925 (2014) doi:10.2134/jeq2014.05.0197 Received 4 May 2014. *Corresponding author ([email protected]).

Landscape Ecology Group, Dep. of Ecology and Environmental Science, Umeå Univ., SE-901 87 Umeå, Sweden. Assigned to Associate Editor Paul DeLaune.

1916

and nutrient availability through more frequent inundation and hyporheic exchange (Xiong and Nilsson, 1997; Spink et al., 1998; Nilsson et al., 2005b). The erosion–deposition activity is anticipated to increase, a broader range of the riparian zone will be temporarily flooded, and the longitudinal variation in flow velocity is expected to increase because of a larger channel complexity (Polvi et al., 2014). Because changes in flow regimes may not be uniform across gradients in stream size, with small streams responding more rapidly to precipitation and snowmelt events, responses to restoration may also vary with stream size. Restoration efforts are commonly evaluated by measuring species diversity and richness of fish, macroinvertebrates, riparian plants, or macrophytes (Palmer et al., 2010; Nilsson et al., 2014). One aspect that is often overlooked, however, is the effect of restoration on soil properties, such as fertility and spatial variability (Spink et al., 1998; Bruland and Richardson, 2005; Unghire et al., 2011). Soil properties are fundamental for the structure and distribution of riparian plant communities, not only directly by controlling plant growth but also indirectly by influencing infiltration properties, water-holding capacity, and nutrient cycling (Helfield et al., 2007; Unghire et al., 2011). Because restored reaches have slower flow and higher retention capacity (Lepori et al., 2005; Engström et al., 2009; Gardeström et al., 2013), riparian soils should accumulate organic and inorganic matter faster following restoration, thus favoring productivity (Naiman et al., 2010). Restoration activities, such as the use of heavy machinery, can also disturb and mix the soil, the effects of which can persevere over long time scales (Bruland and Richardson, 2005). Unghire et al. (2011) found a significant loss of soil organic matter and an increased spatial homogeneity of soils following restoration. Therefore, it should be particularly important to assess the effects of restoration measures on riparian soils, yet few postrestoration studies have prioritized the evaluation of soil properties (Beumer et al., 2008; Unghire et al., 2011; Fournier et al., 2013). In this study, we evaluated restoration effects on riparian soil fertility by using a phytometric assay (Clements and Goldsmith, 1924; Wheeler et al., 1992; Dietrich et al., 2013). Phytometers are standardized plant species used as integrated measures of different environmental conditions (Dietrich et al., 2013). Phytometers have advantages over chemical soil analyses because they directly reveal how a plant experiences its environment (Axmanová et al., 2011). Chemical soil analyses usually determine total, and not plant available, nutrient contents and can thus only serve as proxies for potential habitat productivity. By measuring phytometer performance on soil samples under controlled conditions, the net effect of soil fertility and other soil properties can be assessed (Wheeler et al., 1992; Spink et al., 1998; Axmanová et al., 2011). One major advantage of phytometric assays in the greenhouse is that phytometers will only be limited by soil fertility, whereas several environmental stressors could limit phytometer growth in the field (Wheeler et al., 1992; Spink et al., 1998). Although successful growth in the greenhouse does not guarantee ecosystem health or success of a restoration project in natural vegetation, it may provide indications that can be tested later in the field. Here, we investigated and compared phytometer performance of a forb and a grass species grown on soils extracted from channelized and restored riparian zones of several streams along a climate

gradient and a stream-size gradient in the Vindel River catchment in northern Sweden. We hypothesized that (i) phytometers will perform better in soils from restored than from channelized sites and that (ii) nutrient availability is higher in soils from restored compared to channelized sites due to more frequent flooding and slower flows, depositing organic matter. We also asked (iii) if responses in soil properties to restoration vary with the position in the catchment (e.g., relative to stream size and length of growing season).

Material and Methods Study Sites The Vindel River catchment in northern Sweden (64°00¢– 66°30¢ N, 16°00¢–20°00¢ E; Table 1) drains an area of 12,654 km2 between the Scandes mountain range and the Gulf of Bothnia. The main channel of the catchment is the 455 km long seventhorder Vindel River, a free-flowing river with a flow regime that includes spring flooding after snowmelt, followed by gradually falling water-levels during summer and winter (Renöfält et al., 2005). The dominant vegetation in the catchment is boreal forest, with Norway spruce [Picea abies (L.) H. Karst] and Scots pine (Pinus sylvestris L.) being the most common tree species and ericaceous shrubs (Vaccinium myrtillus L. and V. vitis-idaea L.) dominating in the understory. Slow-weathering gneisses and granites dominate the bedrock material. The main channel and many of its tributaries have been channelized to facilitate timber floating (Nilsson et al., 2005a), but in the last few decades many reaches have been restored by returning coarse sediment (cobbles and boulders) from the banks to the stream or river channel, thus reconnecting the riparian zones and the channel. In this study, we used a paired design of channelized and restored turbulent reaches in a space-for-time substitution approach because no specific prerestoration data were available. Restoration efforts relevant for this study had been conducted between 2002 and 2006. Ten pairs of sites were selected along climate (length of growing season) and stream-size gradients (drainage area), and selection was made so as to avoid correlations between these variables (Fig. 1 and Table 1). We also tried to select pairs of sites that did not differ significantly in any other way than that one site was channelized and the other restored. Soil samples were tested for phytometer response under identical conditions to allow detection of properties of the soil that depend on conditions in the field, such as climate and stream size (a “soil memory”; Schubert et al., 2004). Therefore, we use the terms growing season and stream size even if these variables were not mimicked when phytometer response was tested in the greenhouse. The length of the growing season at the sampling sites (number of days with temperatures greater than +3°C) was used as a proxy for soil activity. It was determined by extrapolation from Ångström (1974) by using values of latitude and altitude, and it varied between 139 and 161 d. The drainage area at every site served as a proxy for stream size (and flooding regime) and was obtained from a Digital Elevation Model (DEM) . The drainage area of sites ranged between 9 and 8202 km2, and sites were grouped into small (1000 km2) stream-size classes (European Commission, 2000; Bejarano et al., 2010). Paired sites (channelized and

www.agronomy.org • www.crops.org • www.soils.org 1917

Table 1. Location of the sites and their position along the climate and stream-size gradients. Pair and status† 1C 1 R (2004) 2C 2 R (2006) 3C 3 R (2002) 4C 4 R (2005) 5C 5 R (2003) 6C 6 R (2003) 7C 7 R (2003) 8C 8 R (2005) 9C 9 R (2004) 10 C 10 R (2005)

Site information Coordinates 65°51¢ N; 17°33¢ E 65°37¢ N; 17°36¢ E 65°45¢ N; 17°15¢ E 65°42¢ N; 17°17¢ E 65°24¢ N; 17°35¢ E 65°31¢ N; 17°08¢ E 65°25¢ N; 17°54¢ E 65°25¢ N; 17°53¢ E 65°21¢ N; 18°04¢ E 65°21¢ N; 17°60¢ E 65°11¢ N; 18°15¢ E 65°10¢ N; 18°15¢ E 65°10¢ N; 18°15¢ E 64°57¢ N; 18°35¢ E 64°55¢ N; 18°46¢ E 64°54¢ N; 18°26¢ E 64°20¢ N; 19°47¢ E 64°23¢ N; 19°29¢ E 64°06¢ N; 20°00¢ E 64°08¢ N; 19°60¢ E

Gradients Altitude

Growing season‡

Stream-size class§

m asl 389 349 414 354 344 422 328 337

d 139.8 143.1 139.0 143.5 145.1 140.1 146.2 145.2

large large medium medium small small medium medium

360 349 282 271 256 240 255 292 229 227 126 191

144.0 145.1 149.6 150.2 150.7 152.8 152.2 151.1 157.4 157.9 161.1 161.1

small small medium medium large large medium small small medium medium medium

† C, channelized; R, restored (year of restoration). ‡ Number of days with temperatures greater than +3C°; extrapolated from Ångström (1974) using latitude and altitude. § Stream-size classes: small, 1000 km2 (European Commission, 2000; Bejarano et al., 2010).

Fig. 1. The 20 experimental sites and their positions in the climate and the stream-size gradients in the Vindel River catchment in northern Sweden. Filled circles are channelized sites (C); open circles are restored sites (R). Numbers refer to the C + R pairs; see Table 1. Stream-size gradient: shaded areas show drainage areas corresponding to every site. Climate gradient: Isotherms (in gray) show length of growing season in days exceeding 3°C (Ångström, 1974; see Table 1 for extrapolation according to altitude at sites). The photographs illustrate typical channelized and restored sites. 1918

Journal of Environmental Quality

restored) were located in the same stream or, where that was not possible, along streams in close vicinity being comparable in size, geomorphology, geology, and climate.

Soil Sampling At all reaches, soil was sampled in the field at the end of June 2009 at two different elevations in the riparian zone. The choice of sampling time (shortly after the spring flood) was motivated by the fact that flooding affects primary production in riparian zones (Dixon and Turner, 2006), and flooding is known to change as a result of this type of restoration (Helfield et al., 2007). The two elevations were determined by dividing the total width of the riparian zone that is exposed to air during the summer into three equal parts and choosing the middle and the upper parts for further study. Riparian zones along streams and rivers in northern Sweden have vegetation that can be divided into belts or zones equivalent to these three elevations, going from a riparian forest community at the top, followed by riparian shrubs and graminoids (Ström et al., 2012). Because all study sites were located along turbulent reaches, the sites that were channelized to make timber floating possible (Nilsson et al., 2005a), the lowest elevation was heavily eroded and basically devoid of fine-grained soil and plants. Therefore, soil sampling was restricted to the two upper elevations. Soil samples were taken from the first 10 cm of topsoil, that is, the biologically most active zone, where the majority of belowground plant biomass is located (Steele et al., 1997), and where impact by erosion is largest (Pinay et al., 1992). Soil material was pooled from three subsamples (each about 3 L) per elevation and site to eliminate microsite variation. Samples were stored in a freezer at -20°C for 7 wk until the beginning of the experiment. This is a common method that is expected to halt soil activity without destroying the geochemical characteristics of the soil microbiota (Ferguson et al., 2003). Therefore, we are confident that any differences in soil fertility between channelized and restored reaches will be revealed even if soil conditions may not perfectly represent the riparian environment.

Soil Analysis About half the volume of each soil sample was dried at 105°C for 72 h to constant mass. The organic matter content was estimated by loss-on-ignition at 450°C for 8 h. Soil fractions were determined after removal of organic matter (i.e., particle sizes 16, 8, 4, 2, 1, 0.5, 0.25, 0.125, 0.075 mm diameter, respectively). The soil water-holding capacity (j) was measured indirectly by calculating substrate fineness by weighting log-transformed values of mean particle size by percentage composition of substrate based on three j values according to Wentworth: clay and silt (9.0 to 6.5), sand (6.5 to 2.0), and gravel (2.0 to -2.0) (Nilsson et al., 1989). The mass fractions of carbon (C) and nitrogen (N) along with the proportions of the stable isotopes 13C and 15N were analyzed using elemental analyzer–isotope ratio mass spectrometer (Flash EA 2000 and DeltaV, both Thermo Fisher Scientific). The dry mass was estimated by weighing samples after oven drying at 70°C for a minimum of 18 h. The C and N of the dried sample material were converted to CO2 and N2 by combustion, measured mass spectrometrically, and reported as mass fractions (g C or N per g dry mass). The isotopic ratios were expressed as d13C = 13C/12C,

using Vienna PeeDee Belemnite-C (VPDB scale; available from the International Atomic Energy Agency), and as d15N = 15N/14N, using the atmospheric N2 scale. Working standards were calibrated against reference standards. The results were corrected for drift and sample size effect (nonlinearity) (Ohlsson and Wallmark, 1999). Due to the noncalcareous bedrock in the area, total C can be considered to be equal to organic C.

Phytometer Experiment A phytometric assay using the forb Centaurea cyanus L. (cornflower), Asteraceae, and the grass Phleum pratense L. (timothy), Poaceae, was conducted in a greenhouse at Umeå University, Umeå, Sweden, between 17 Aug. 2009 and 18 Jan. 2010. We used these species, which are native to Sweden, to cover the two most common forms of plant species and for having seeds with little endosperm to ensure their dependence on soil quality. Also, we used commercial seeds to ensure more constant seed material. The upper parts of the riparian zone, in which the present study was located, do not have any specific, exclusively riparian species. Instead, the vast majority of species in the region, including meadow plants as the ones used in this study, can be found there. Three seeds of each species were planted in four 300-mL pots (250 mL of soil per pot) of each of the 40 different soils (10 + 10 sites, 2 elevations). After germination, the seedlings in each pot were thinned to retain the most vital plant. Temperature (20°C), water availability and light regime were kept constant, keeping the pots constantly moist and providing 12 h of daylight plus artificial light per day after the autumnal equinox. All senesced material was collected to be included in the final biomass measurement and the number of produced flowers was recorded during the experimental period. After 21 wk, plant height was measured and aboveground biomass determined after oven drying at 35°C to constant weight.

Data Analysis Data analysis was performed using IBM SPSS version 20 and the statistical package R (R Development Core Team, 2012). Shapiro-Wilk’s test was used to test for normality, and data were transformed to better fit assumptions of normality when necessary (log10 for phytometer biomass and number of flowers, soil C and N contents, as well as contents of gravel and organic matter; arcsine square root for clay–silt content). To determine if restoration affected phytometer performance, nutrient levels, or soil properties, we used two-way mixedeffects analysis of variance (ANOVA) in SPSS, distinguishing between channelized and restored sites (status as fixed and pair as random factor, i.e., blocking variable) in the three streamsize classes. All further analyses were performed with data from channelized and restored sites separately to be able to detect differing trends that point to changes in ecological processes attributable to restoration. We used Pearson’s correlation analysis to detect trends in the phytometer responses (P. pratense biomass; C. cyanus biomass and number of flowers) related to the abiotic variables (j, as a collective value for the soil fractions; organic matter content; contents of soil C and N; proportions of the stable isotopes d13C and d15N). The effect of stream size was determined using one-way ANOVA combined with a Tukey’s post hoc test. Trends in phytometer responses or abiotic

www.agronomy.org • www.crops.org • www.soils.org 1919

variables related to the length of growing season were detected using Pearson’s correlation analysis, and regression (linear and quadratic) was used to generate trend lines with the best fit to represent the data.

Results Phytometer Response Variables Even though we were unable to detect an overall effect of restoration irrespective of stream size (P > 0.05, two-way ANOVA), phytometer performance was consistently better in soils from restored than from channelized sites along small streams (P. pratense: biomass P = 0.034; C. cyanus: biomass P = 0.005, number of flowers P = 0.014; two-way ANOVA; Fig. 2a). A similar, but nonsignificant trend was observed for soil from medium-sized streams. For soil from large streams, however, the trend was reversed, with better, although not significant, phytometer performance at channelized reaches. The number

of C. cyanus flowers was correlated with stream size at restored sites, with highest values in soil from medium-sized streams (Fig. 2a; P < 0.05, one-way ANOVA). We did not, however, observe such patterns for the other phytometer responses or in soils from channelized sites. Phytometer performance showed a generally increasing trend with soils from reaches with increasing length of growing season (Fig. 2b). This trend was, however, only statistically significant at channelized sites and differed between P. pratense, which showed a linear increase and C. cyanus, which showed an overall increase, although with a quadratic relationship showing best fit, with lowest values at the medium-length growing season. The number of C. cyanus flowers did not vary significantly with soils from reaches with different lengths of growing season (P > 0.05). The phytometer performance generally decreased with increasing proportions of the stable isotopes 13C and 15N (Table 2). For soils from restored sites, the phytometers performed better with increasing values of soil water-holding capacity (j), and

Fig. 2. Responses of phytometers when grown on soils collected at channelized and restored sites. (a) Boxplot showing differences among streamsize classes (small, medium, large) and restoration status. Significant effects of restoration are indicated as follows: *P < 0.05, **P < 0.01 (two-way ANOVA; restoration status as fixed and pair as random factors); effects of stream-size classes are indicated by different letters (one-way ANOVA; significant at the 0.05 level). (b) Regressions with length of growing season. Best fit lines (R2) are shown if significant for C, channelized, and R, restored, sites (levels of significance: *P < 0.05, **P < 0.01). Note different scales. 1920

Journal of Environmental Quality

Table 2. Pearson’s correlation coefficients for phytometer responses and abiotic variables for Phleum pratense and Centaurea cyanus at channelized and restored sites (n = 20 each). Only significant correlations are shown. Phytometer response variables Channelized sites P. pratense biomass C. cyanus biomass Water-holding capacity Organic matter Carbon Nitrogen C/N d13C d15N

C. cyanus flowers

– – – – –

– – – – –

– – – – –

–0.500* –

– –0.533*

– –0.496*

Restored sites P. pratense biomass C. cyanus biomass – 0.683** 0.525* 0.476* 0.599** –0.466* –0.496*

C. cyanus flowers

0.595** 0.588** 0.455* – 0.543*

– – – – 0.553*

–0.566** –0.620**

–0.532* –

* Correlation is significant at the 0.05 level (two-tailed). ** Correlation is significant at the 0.01 level (two-tailed).

contents of organic matter, soil C, soil N, and soil C/N. In soils from channelized sites, we observed few significant correlations between phytometer performance and environmental variables, but P. pratense showed a negative correlation with soil d13C, whereas C. cyanus responses were negatively correlated with soil d15N. In contrast, in soils from restored sites, we found differences in responses between species with respect to soil N (significant only for P. pratense) and j (significant only for C. cyanus). Furthermore, the number of C. cyanus flowers was only significantly related to the C/N ratio and d13C, whereas C. cyanus biomass related to all abiotic variables except soil N.

Abiotic Variables When soil sampling was conducted, the water levels were at their summer low and the riparian vegetation had developed. Because of this, and because our study streams are supply-limited in terms of fine sediments (Rosenfeld et al., 2011), we could not see any clear signs of recent erosion or deposition events at most sites. The soil content of C and N indicated no effects of either restoration or stream size, but the range of values decreased with increasing stream size (Fig. 3a, P > 0.05, one-way ANOVA). The C/N ratio decreased significantly with increasing stream size but only in soils from restored sites. The proportion of d15N was significantly higher in soils from restored sites in small streams (P = 0.006). Furthermore, in soils from restored sites, the proportion of d15N increased significantly with increasing stream size (Fig. 3a, P < 0.05, one-way ANOVA), whereas no such trend was observed in soils from channelized sites. Also, the proportion of d13C was related to stream size in soils from restored sites, but not in soils from channelized sites, with the lowest values in soils from medium-sized streams and the highest values in soils from large streams. Both soil C and N contents were quadratically related to the length of the growing season at restored sites, with the lowest values at the medium-length growing season (Fig. 3b). The level of d13C decreased linearly with an increasing growing season in soils from channelized sites only. The proportion of d15N showed a linear decrease in soils from both channelized and restored sites. The C/N ratio was not significantly related to the length of the growing season. Overall, the proportions of different inorganic soil fractions and organic matter did not differ significantly between channelized and restored sites (P > 0.05, two-way ANOVA; Fig.

4). When the stream size increased, the proportion of organic matter showed a decreasing trend and the proportion of gravel an increasing trend in soils from both channelized and restored streams, but the patterns were not statistically significant (P > 0.05, one-way ANOVA). The proportions of clay–silt and sand were more or less constant among the stream-size classes. We observed subtle differences between soils from channelized and restored sites in small streams, with more organic matter and less gravel available in soils from restored reaches. The soil water-holding capacity (j), which is a function of soil grain-size distribution, seemed to decrease with increasing stream size in soils from restored reaches, whereas in soils from channelized sites, a quadratic relationship between j and stream size best described the data, with highest values in soils from the medium-sized streams (R2 = 0.352; p = 0.025). At both channelized and restored sites, soils from the large streams appeared to have the lowest water-holding capacity (Fig. 4). Also, the variation among sites increased with increasing stream size.

Discussion This phytometer study was designed to evaluate whether restoration of channelized stream reaches affects the quality of riparian soils. We found that the phytometers grew significantly better in the soils from restored sites in small streams. The higher dry weight of phytometers at the end of the experiment reflects a potentially higher productivity of these soils and suggests a restoration effect (Al-Farray et al., 1984). The similar patterns for medium-sized, but opposite trends for large streams, could mean that recovery was more efficient in small streams but could also indicate that the recovery just took longer when stream size increased. Small streams may recover more rapidly after restoration because they experience a more variable flood regime with more frequent but shorter and less severe flooding events following restoration (Helfield et al., 2007; Burt et al., 2010). This induces a shift in dominance from dwarf shrubs to grasses and forbs (Helfield et al., 2007), the latter producing more litter and decomposing faster (Olofsson and Oksanen, 2002), which should enhance soil productivity. In addition, the more variable flood regime means that there are repeated events of sediment and nutrient deposition that can favor plants (Spink et al., 1998; Nilsson et al., 2005b). In contrast, large rivers experience longer floods with higher stream power that may inhibit or delay soil recovery and plant growth after restoration (Burt et al., 2010).

www.agronomy.org • www.crops.org • www.soils.org 1921

Fig. 3. Soil carbon and nitrogen contents and the amounts of their isotopes at channelized and restored sites. (a) Boxplot showing differences among stream-size classes (small, medium, large). Significant effects of restoration are indicated as follows: *P < 0.05, **P < 0.01 (two-way ANOVA; restoration status as fixed and pair as random factors); effects of stream-size classes are indicated by different letters (one-way ANOVA; significant at the 0.05 level). (b) Regressions with length of growing season. Best fit lines (R2) are shown if significant for C, channelized, and R, restored, sites (levels of significance: *P < 0.05, **P < 0.01). Note different scales.

1922

Journal of Environmental Quality

Fig. 4. Proportions of different soil fractions and organic (org.) matter in soils collected at channelized (C) and restored (R) sites in the different stream-size classes. The j-value above the columns indicates the accumulated water-holding capacity of the mineral soil.

The large variation in soil C and N data in small as compared to medium and large streams may reflect these differences (Fig. 3a). It is also likely that such differences between stream sizes explain why we did not observe an overall direct effect of restoration in the entire catchment. Dunn et al. (2011) found that 99% of a river network in Kansas, USA, is first- to third-order streams. Similar numbers have been presented for Sweden, with catchment areas smaller than 15 km2 in >90% of the total channel length (Bishop et al., 2008). The drainage areas in the small stream-size class (10 yr. In our study, we detected an effect on soil conditions relevant to vegetation in small streams via the phytometers already 3 to 7 yr after restoration. These results and the above-mentioned trajectory of recovery lead us to suggest that responses will be gradual and begin to occur sooner than 10 yr after restoration in small streams. With longer growing seasons, soil processes will act over longer time periods, which may explain the observed increase in phytometer performance with soils from reaches with increasing length of growing season. However, trends associated with length of growing season were only statistically significant at channelized sites (Fig. 2 and Table 2). This could possibly be the result of the soils at these sites having stabilized during the more than one-centurylong period of active timber floating (Törnlund and Östlund, 2002; Nilsson et al., 2005a). In contrast, soils at restored sites are likely still affected by, and have not yet recovered from, the disturbance of restoration (Tullos et al., 2009; Catford et al., 2012). Therefore, soils from restored sites may not yet be affected by the length of growing season but rather by the local variability of environmental factors, as reflected by the large variation in soil data from these sites. Helfield et al. (2007) investigated stream reaches that had been restored in the same way as those in the present study and found that flooding frequency was higher in restored than in channelized reaches. They emphasized the role of changes in flood disturbance to explain increases in riparian plant species richness and cover within 3 to 10 yr of restoration. This study shows that increases in soil fertility, which favor more species-rich forb and grass communities, may also contribute to postrestoration

www.agronomy.org • www.crops.org • www.soils.org 1923

vegetation responses. However, we cannot assert to what extent soil changes drive vegetation responses, and to what extent soil changes are caused by changes in vegetation driven by changes in flooding and flow velocity. We found no effect of water-holding capacity on phytometer performance, which may be explained by the watering scheme in the greenhouse, where pots were not allowed to dry out. Like Beumer et al. (2008), we observed species-specific phytometer responses. For example, in soils from channelized sites, the aboveground biomass of P. pratense was correlated with the proportion of d13C in the soil, whereas the C. cyanus performance was more strongly related to the availability of d15N. This result points to the advantage of using a combination of species with different requirements, for example, from different growth forms of plant species, to cover as much as possible of the variation in soil characteristics (cf. Johansson and Nilsson, 2002). Furthermore, the choice of phytometer species should be adjusted to suit the environmental conditions, such as soil pH (Axmanová et al., 2011).

Management Implications The use of phytometers in greenhouse experiments is a fast and easy approach for assessing soil quality. The major advantage compared with chemical analyses is that phytometers specifically record soil attributes relevant for plant performance. In comparison to field experiments, soil fertility estimates will not be masked by other environmental stressors affecting plants and therefore can provide reliable information about soil conditions. However, a limitation of our study is that the soil samples do not record the seasonal variation in soil properties or nutrient availability year-round. We found a positive effect of restoration on phytometers grown in soils from small streams. Therefore, we suggest that it is best to initiate restoration efforts in small streams where soils appear to recover faster. Given that the proportion of low-order streams is high and that positive effects of restoration in headwaters may also affect downstream reaches, initiating restoration activities in tributaries may favorably influence the response rate of the ecosystem in larger streams.

Acknowledgments We thank Eliza Maher Hasselquist and Niles Hasselquist for valuable input concerning the isotope data. Comments from two journal reviewers greatly improved the manuscript. Funding was provided by the Swedish research council Formas (to CN).

References Al-Farray, M.M., K.E. Giller, and B.D. Wheeler. 1984. Phytometric assessment of fertility of waterlogged rich-fen peats using Epilobium hirsutum L. Plant Soil 81:283 –289. Ångström, A. 1974. Sveriges klimat. 3rd ed. Generalstabens Litografiska Anstalts Förlag, Stockholm. Axmanová, I., D. Zelený, C.-F. Li, and M. Chytrý. 2011. Environmental factors influencing herb layer productivity in central European oak forests: Insights from soil and biomass analyses and a phytometer experiment. Plant Soil 342:183–194. doi:10.1007/s11104-010-0683-9 Baxter, C.V., K.D. Fausch, and W.C. Saunders. 2005. Tangled webs: Reciprocal flows of invertebrate prey link streams and riparian zones. Freshwater Biol. 50:201–220. doi:10.1111/j.1365-2427.2004.01328.x Bejarano, M.D., M. Marchamalo, and D. Garcia de Jalon. 2010. Flow regime patterns and their controlling factors in the Ebro basin (Spain). J. Hydrol. 385:323–335. doi:10.1016/j.jhydrol.2010.03.001 Bernhardt, E.S., M.A. Palmer, J.D. Allan, G. Alexander, K. Barnas, S. Brooks, J. Carr, S. Clayton, C. Dahm, J. Follstad-Shah, D. Galat, S. Gloss, P. Goodwin, D. Hart, B. Hassett, R. Jenkinson, S. Katz, G.M. Kondolf, P.S. 1924

Lake, R. Lave, J.L. Meyer, T.K. O’Donnell, L. Pagano, B. Powell, and S. Sudduth. 2005. Ecology: Synthesizing US river restoration efforts. Science 308:636–637. doi:10.1126/science.1109769 Beumer, V., J.N. Ohm, G. van Wirdum, B. Beltman, J. Griffioen, and J.T.A. Verhoeven. 2008. Biogeochemical plant site conditions in stream valleys after winter flooding: A phytometer approach. Biogeosciences Discuss. 5:5203–5232. doi:10.5194/bgd-5-5203-2008 Bishop, K., I. Buffam, M. Erlandsson, J. Fölster, H. Laudon, J. Seibert, and J. Temnerud. 2008. Aqua Incognita: The unknown headwaters. Hydrol. Processes 22:1239–1242. doi:10.1002/hyp.7049 Boulton, A.J., S. Findlay, P. Marmonier, E.H. Stanley, and H.M. Valett. 1998. The functional significance of the hyporheic zone in streams and rivers. Annu. Rev. Ecol. Syst. 29:59–81. doi:10.1146/annurev.ecolsys.29.1.59 Bruland, G.L., and C.J. Richardson. 2005. Spatial variability of soil properties in created, restored, and paired natural wetlands. Soil Sci. Soc. Am. J. 69:273–284. Burt, T., G. Pinay, and S. Sabater. 2010. What do we still need to know about the ecohydrology of riparian zones? Ecohydrology 3:373–377. doi:10.1002/ eco.140 Catford, J.A., C.C. Daehler, H.T. Murphy, A.W. Sheppard, B.D. Hardesty, D.A. Westcott, M. Rejmánek, P.J. Bellingham, J. Pergl, C.C. Horvitz, and P.E. Hulme. 2012. The intermediate disturbance hypothesis and plant invasions: Implications for species richness and management. Perspect. Plant Ecol. Evol. Syst. 14:231–241. doi:10.1016/j.ppees.2011.12.002 Clements, F.E., and G.W. Goldsmith. 1924. The phytometer method in ecology: The plant and community as instruments. Carnegie Institution, Washington, DC. Dawson, T.E., S. Mambelli, A.H. Plamboeck, P.H. Templer, and K.P. Tu. 2002. Stable isotopes in ecology. Annu. Rev. Ecol. Syst. 33:507–559. doi:10.1146/annurev.ecolsys.33.020602.095451 Dietrich, A.L., C. Nilsson, and R. Jansson. 2013. Phytometers are underutilised for evaluating ecological restoration. Basic Appl. Ecol. 14:369–377. doi:10.1016/j.baae.2013.05.008 Dixon, M.D., and M.G. Turner. 2006. Simulated recruitment of riparian trees and shrubs under natural and regulated flow regimes on the Wisconsin River, USA. River Res. Appl. 22:1057–1083. doi:10.1002/rra.948 Dudgeon, D., A.H. Arthington, M.O. Gessner, Z.I. Kawabata, D.J. Knowler, C. Leveque, R.J. Naiman, A.H. Prieur-Richard, D. Soto, M.L.J. Stiassny, and C.A. Sullivan. 2006. Freshwater biodiversity: Importance, threats, status and conservation challenges. Biol. Rev. Camb. Philos. Soc. 81:163–182. doi:10.1017/S1464793105006950 Dunn, W.C., B.T. Milne, R. Mantilla, and V.K. Gupta. 2011. Scaling relations between riparian vegetation and stream order in the Whitewater River network, Kansas, USA. Landscape Ecol. 26:983–997. doi:10.1007/ s10980-011-9622-2 European Commission. 2000. Directive 2000/60/EC of the European Parliament and of the Council of October the 23rd establishing a framework for community action in the field of water policy. Official Journal, L 327, 22/12/2000, p. 1–73. Engström, J., C. Nilsson, and R. Jansson. 2009. Effects of stream restoration on dispersal of plant propagules. J. Appl. Ecol. 46:397–405. doi:10.1111/j.1365-2664.2009.01612.x Ferguson, S.H., P.D. Franzmann, I. Snape, A.T. Revill, M.G. Trefry, and L.R. Zappia. 2003. Effects of temperature on mineralisation of petroleum in contaminated Antarctic terrestrial sediments. Chemosphere 52:975–987. doi:10.1016/S0045-6535(03)00265-0 Fournier, B., C. Guenat, G. Bullinger-Weber, and E.A.D. Mitchell. 2013. Spatiotemporal heterogeneity of riparian soil morphology in a restored floodplain. Hydrol. Earth Syst. Sci. 17:4031–4042. doi:10.5194/hess-17-4031-2013 Gardeström, J., D. Holmqvist, L.E. Polvi, and C. Nilsson. 2013. Demonstration restoration measures in tributaries of the Vindel river catchment. Ecol. Soc. 18(3):8. doi:10.5751/ES-05609-180308 Gift, D.M., P.M. Groffman, S.S. Kaushal, and P.M. Mayer. 2010. Denitrification potential, root biomass, and organic matter in degraded and restored urban riparian zones. Restor. Ecol. 18:113–120. doi:10.1111/j.1526-100X.2008.00438.x Haycock, N.E., G. Pinay, and C. Walker. 1993. Nitrogen-retention in river corridors: European perspective. Ambio 22:340–346. Helfield, J.M., S.J. Capon, C. Nilsson, R. Jansson, and D. Palm. 2007. Restoration of rivers used for timber floating: Effects on riparian plant diversity. Ecol. Appl. 17:840–851. doi:10.1890/06-0343 Hill, A.R. 1996. Nitrate removal in stream riparian zones. J. Environ. Qual. 25:743–755. doi:10.2134/jeq1996.00472425002500040014x Högberg, P. 1997. 15N natural abundance in soil-plant systems. New Phytol. 137:179–203. doi:10.1046/j.1469-8137.1997.00808.x Janssen, B.H. 1996. Nitrogen mineralization in relation to C:N ratio and Journal of Environmental Quality

decomposability of organic materials. Plant Soil 181:39–45. doi:10.1007/ BF00011290 Johansson, M.E., and C. Nilsson. 2002. Responses of riparian plants to flood variables along free-flowing and regulated rivers. J. Appl. Ecol. 39:971– 986. doi:10.1046/j.1365-2664.2002.00770.x Lepori, F., D. Palm, and B. Malmqvist. 2005. Effects of stream restoration on ecosystem functioning: Detritus retentiveness and decomposition. J. Appl. Ecol. 42:228–238. doi:10.1111/j.1365-2664.2004.00965.x Lowrance, R. 1992. Nitrogen inputs from a field-sized agricultural watershed. J. Environ. Qual. 21:602–607. doi:10.2134/ jeq1992.00472425002100040013x Maddock, I. 1999. The importance of physical habitat assessment for evaluating river health. Freshwater Biol. 41:373–391. doi:10.1046/j.1365-2427.1999.00437.x Malmqvist, B., and S. Rundle. 2002. Threats to the running water ecosystems of the world. Environ. Conserv. 29:134–153. doi:10.1017/ S0376892902000097 McClain, M.E., E.W. Boyer, C.L. Dent, S.E. Gergel, N.B. Grimm, P.M. Groffman, S.C. Hart, J.W. Harvey, C.A. Johnston, E. Mayorga, W.H. McDowell, and G. Pinay. 2003. Biogeochemical hot spots and hot moments at the interface of terrestrial and aquatic ecosystems. Ecosystems 6:301–312. doi:10.1007/ s10021-003-0161-9 Meyer, J.L., D.L. Strayer, J.B. Wallace, S.L. Eggert, G.S. Helfman, and N.E. Leonard. 2007. The contribution of headwater streams to biodiversity in river networks. J. Am. Water Resour. Assoc. 43:86–103. doi:10.1111/j.1752-1688.2007.00008.x Muotka, T., and J. Syrjänen. 2007. Changes in habitat structure, benthic invertebrate diversity, trout populations and ecosystem processes in restored forest streams: A boreal perspective. Freshwater Biol. 52:724– 737. doi:10.1111/j.1365-2427.2007.01727.x Naiman, R.J., J.S. Bechtold, T.J. Beechie, J.J. Latterell, and R. Van Pelt. 2010. A process-based view of floodplain forest patterns in coastal river valleys of the Pacific Northwest. Ecosystems 13:1–31. doi:10.1007/s10021-009-9298-5 Naiman, R.J., and H. Décamps. 1997. The ecology of interfaces: Riparian zones. Ann. Rev. Ecol. Syst. 28:621‒658. Naiman, R.J., H. Décamps, and M. Pollock. 1993. The role of riparian corridors in maintaining regional biodiversity. Ecol. Appl. 3:209–212. doi:10.2307/1941822 Naiman, R.J., K.L. Fetherston, S. McKay, and J. Chen. 1998. Riparian forests. In: R.J. Naiman and R.E. Bilby, editors, River ecology and management: Lessons from the Pacific Coastal region. Springer, New York. p. 289–323. Nilsson, C., A. Ekblad, M. Dynesius, S. Backe, M. Gardfjell, B. Carlberg, S. Hellqvist, and R. Jansson. 1994. Comparison of species richness and traits of riparian plants between a main river channel and its tributaries. J. Ecol. 82:281‒295. Nilsson, C., G. Grelsson, M. Johansson, and U. Sperens. 1989. Patterns of plant species richness along riverbanks. Ecology 70:77–84. doi:10.2307/1938414 Nilsson, C., F. Lepori, B. Malmqvist, E. Törnlund, N. Hjerdt, J.M. Helfield, D. Palm, J. Östergren, R. Jansson, E. Brännäs, and H. Lundqvist. 2005a. Forecasting environmental responses to restoration of rivers used as log floatways: An interdisciplinary challenge. Ecosystems 8:779–800. doi:10.1007/s10021-005-0030-9 Nilsson, C., L.E. Polvi, J. Gardeström, E.M. Hasselquist, L. Lind, and J.M. Sarneel. 2014. Riparian and in-stream restoration of boreal streams and rivers: Success or failure? Ecohydrology (in press). Nilsson, C., C.A. Reidy, M. Dynesius, and C. Revenga. 2005b. Fragmentation and flow regulation of the world’s large river systems. Science 308:405– 408. doi:10.1126/science.1107887 Ohlsson, K.E.A., and P.H. Wallmark. 1999. Novel calibration with correction for drift and non-linear response for continuous flow isotope ratio mass spectrometry applied to the determination of d15N, total nitrogen, d13C and total carbon in biological material. Analyst (Lond.) 124:571–577. doi:10.1039/a900855a Olofsson, J., and L. Oksanen. 2002. Role of litter decomposition for the increased primary productivity in areas heavily grazed by reindeer: A litterbag experiment. Oikos 96:507–515. doi:10.1034/j.1600-0706.2002.960312.x Palmer, M.A., E.S. Bernhardt, J.D. Allan, P.S. Lake, G. Alexander, S. Brooks, J. Carr, S. Clayton, C.N. Dahm, J.F. Shah, D.L. Galat, S.G. Loss, P. Goodwin, D.D. Hart, B. Hassett, R. Jenkinson, G.M. Kondolf, R. Lave, J.L. Meyer, T.K. O’Donnell, L. Pagano, and E. Sudduth. 2005. Standards for ecologically successful river restoration. J. Appl. Ecol. 42:208–217. doi:10.1111/j.1365-2664.2005.01004.x Palmer, M.A., H.L. Menninger, and E. Bernhardt. 2010. River restoration,

habitat heterogeneity and biodiversity: A failure of theory or practice? Freshwater Biol. 55:205–222. doi:10.1111/j.1365-2427.2009.02372.x Pinay, G., A. Fabre, P. Vervier, and F. Gazelle. 1992. Control of C, N, P distribution in soils of riparian forests. Landscape Ecol. 6:121–132. doi:10.1007/BF00130025 Polvi, L.E., C. Nilsson, and E.M. Hasselquist. 2014. Potential and actual geomorphic complexity of restored headwater streams in northern Sweden. Geomorphology 210:98–118. doi:10.1016/j.geomorph.2013.12.025 R Development Core Team. 2012. R: A language and environment for statistical computing. R version 2.15.2. The R Foundation for Statistical Computing, Vienna, Austria. Renöfält, B., R. Jansson, and C. Nilsson. 2005. Spatial patterns of plant invasiveness in a riparian corridor. Landscape Ecol. 20:165–176. doi:10.1007/s10980-004-2262-z Roni, P., K. Hanson, and T. Beechie. 2008. Global review of the physical and biological effectiveness of stream habitat rehabilitation techniques. N. Am. J. Fish. Manage. 28:856–890. doi:10.1577/M06-169.1 Rosenfeld, J.S., D. Hogan, D. Palm, H. Lundqvist, and C. Nilsson. 2011. Contrasting landscape influences on sediment supply and stream restoration trajectories in northern Fennoscandia (Sweden and Finland) and coastal British Columbia. Environ. Manage. 47:28–39. doi:10.1007/ s00267-010-9585-0 Schubert, S.D., M.J. Suarez, P.J. Pegion, R.D. Koster, and J.T. Bacmeister. 2004. Causes of long-term drought in the US Great Plains. J. Clim. 17:485–503. doi:10.1175/1520-0442(2004)0172.0.CO;2 SER. 2004. The SER international primer on ecological restoration. Society for Ecological Restoration International, Science and Policy Working Group. http://www.ser.org/resources/resources-detail-view/ser-internationalprimer-on-ecological-restoration (accessed 16 Sept. 2014). Spink, A., R.E. Sparks, M. van Oorschot, and J.T.A. Verhoeven. 1998. Nutrient dynamics of large river floodplains. Regul. Rivers Res. Manage. 14:203–216. doi:10.1002/ (SICI)1099-1646(199803/04)14:23.0.CO;2-7 Steele, S.J., S.T. Gower, J.G. Vogel, and J.M. Norman. 1997. Root mass, net primary production and turnover in aspen, jack pine and black spruce forests in Saskatchewan and Manitoba, Canada. Tree Physiol. 17:577– 587. doi:10.1093/treephys/17.8-9.577 Ström, L., R. Jansson, and C. Nilsson. 2012. Projected changes in plant species richness and extent of riparian vegetation belts as a result of climate-driven hydrological change along the Vindel River in Sweden. Freshwater Biol. 57:49–60. doi:10.1111/j.1365-2427.2011.02694.x Thorp, J.H., M.C. Thoms, and M.D. Delong. 2006. The riverine ecosystem synthesis: Biocomplexity in river networks across space and time. River Res. Appl. 22:123–147. doi:10.1002/rra.901 Törnlund, E., and L. Östlund. 2002. Floating of timber in northern Sweden: The construction of floatways and transformation of rivers. Environ. Hist. 8:85–106. doi:10.3197/096734002129342611 Tullos, D.D., D.L. Penrose, G.D. Jennings, and W.G. Cope. 2009. Analysis of functional traits in reconfigured channels: Implications for the bioassessment and disturbance of river restoration. J. North Am. Benthol. Soc. 28:80–92. doi:10.1899/07-122.1 Unghire, J.M., A.E. Sutton-Grier, N.E. Flanagan, and C.J. Richardson. 2011. Spatial impacts of stream and wetland restoration on riparian soil properties in the North Carolina Piedmont. Restor. Ecol. 19:738–746. doi:10.1111/j.1526-100X.2010.00726.x Vannote, R.L., G.W. Minshall, K.W. Cummins, J.R. Sedell, and C.E. Cushing. 1980. The river continuum concept. Can. J. Fish. Aquat. Sci. 37:130–137. doi:10.1139/f80-017 Ward, J.V. 1989. The 4-dimensional nature of lotic ecosystems. J. North Am. Benthol. Soc. 8:2–8. doi:10.2307/1467397 Ward, J.V., K. Tockner, D.B. Arscott, and C. Claret. 2002. Riverine landscape diversity. Freshwater Biol. 47:517–539. doi:10.1046/j.1365-2427.2002.00893.x Wheeler, B.D., S.C. Shaw, and R.E.D. Cook. 1992. Phytometric assessment of the fertility of undrained rich-fen soils. J. Appl. Ecol. 29:466–475. doi:10.2307/2404514 Xiong, S., and C. Nilsson. 1997. Dynamics of leaf litter accumulation and its effects on riparian vegetation: A review. Bot. Rev. 63:240–264. doi:10.1007/BF02857951

www.agronomy.org • www.crops.org • www.soils.org 1925