Content uploaded by Steve J. Ormerod
Author content
All content in this area was uploaded by Steve J. Ormerod on Feb 11, 2018
Content may be subject to copyright.
Accepted Article
This article has been accepted for publication and undergone full peer review but has not been
through the copyediting, typesetting, pagination and proofreading process, which may lead to
differences between this version and the Version of Record. Please cite this article as doi:
10.1111/gcb.13103
This article is protected by copyright. All rights reserved.
Received Date : 24-Feb-2015
Accepted Date : 21-Aug-2015
Article type : Primary Research Articles
Beyond cool: adapting upland streams for climate change using riparian woodlands
STEPHEN M. THOMAS
1,2
*, SIÂN W. GRIFFITHS
1
and S. J. ORMEROD
1
1. Catchment Research Group, Cardiff School of Biosciences, Cardiff University, Sir Martin
Evans Building, Museum Avenue, Cardiff, CF10 3AX, UK.
2. Department of Environmental Sciences, University of Helsinki, P.O. Box 65, FI-00014,
Finland
*Correspondence: Department of Fish Ecology and Evolution, EAWAG Swiss Federal Institute
of Aquatic Science and Technology, Center for Ecology, Evolution and Biogeochemistry,
Seestrasse 79, CH-6047 Kastanienbaum, Switzerland. Email: stephen.thomas@eawag.ch
Running head: Riparian woodlands and climate change adaptation
Keywords: Adaptation, buffer strip, CPOM, isotope, macroinvertebrate, river, subsidy
Accepted Article
This article is protected by copyright. All rights reserved.
Type of Paper: Primary Research Article
Abstract
Managed adaptation could reduce the risks of climate change to the world’s ecosystems, but there
have been surprisingly few practical evaluations of the options available. For example, riparian
woodland is advocated widely as shade to reduce warming in temperate streams, but few studies
have considered collateral effects on species composition or ecosystem functions. Here, we use
cross sectional analyses at two scales (region and within streams) to investigate whether four
types of riparian management, including those proposed to reduce potential climate change
impacts, might also affect the composition, functional character, dynamics and energetic
resourcing of macroinvertebrates in upland Welsh streams (UK).
Riparian land use across the region had only small effects on invertebrate taxonomic composition,
while stable isotope data showed how energetic resources assimilated by macroinvertebrates in all
functional guilds were split roughly 50:50 between terrestrial and aquatic origins irrespective of
riparian management. Nevertheless, streams draining the most extensive deciduous woodland had
the greatest stocks of coarse particulate matter (CPOM) and greater numbers of “shredding”
detritivores. Stream-scale investigations showed that macroinvertebrate biomass in deciduous
woodland streams was around twice that in moorland streams, and lowest of all in streams
draining non-native conifers.
Accepted Article
This article is protected by copyright. All rights reserved.
The unexpected absence of contrasting terrestrial signals in the isotopic data implies that factors
other than local land use affect the relative incorporation of allochthonous subsidies into riverine
food webs. Nevertheless, our results reveal how planting deciduous riparian trees along temperate
headwaters as an adaptation to climate change can modify macroinvertebrate function, increase
biomass and potentially enhance resilience by increasing basal resources where cover is extensive
(>60m riparian width). We advocate greater urgency in efforts to understand the ecosystem
consequences of climate change adaptation in order to guide future actions.
Introduction
Although reducing greenhouse gas emissions is fundamental to mitigating future climate change,
there is growing expectation that further increase in global temperature cannot now be avoided
(IPCC, 2014). Interest is growing, therefore, in strategies for climate change adaptation that
might minimize the worst effects on key resources (Perry, 2015). These include organisms,
ecosystems and the many services that they provide, and as a result ecologists have been among
the strongest advocates for climate change adaptation (Hulme, 2005; Dudgeon et al., 2006; Seavy
et al., 2009). The broad aims involve predicting the effects on vulnerable species or habitats,
increasing their resilience, maintaining sensitive species or assemblages, restoring lost
connectivity, reducing the stressors with which climate change interacts, and providing security
for critical ecosystems (Hulme, 2005; Ormerod, 2009; Palmer et al. 2009; Seavy et al., 2009). So
far, however, there are few specific examples where approaches advocated in theory have been
evaluated in practice (Mawdsley et al., 2009; Macgregor & van Dijk, 2014). This is an important
knowledge gap given the extent of actions likely to be required to adapt ecosystems to climate
change, and because of the potentially far-reaching effects on the risks, benefits and services that
might arise.
Accepted Article
This article is protected by copyright. All rights reserved.
Stream and river ecosystems have figured strongly in the adaptation debate for two major
reasons. First, they have major global value to human life support, for example through water
supply, flood regulation, pollutant disposal, support for major biogeochemical cycles, and critical
roles in fisheries (Holmlund & Hammer, 1999; Wilson & Carpenter, 1999; Ormerod, 2009).
Secondly, streams and rivers are among the most sensitive of all ecosystems to climate change
because they are coupled directly to the global hydrological cycle, linked closely to atmospheric
thermal regimes, and at risk from interactions between climate change and existing anthropogenic
stressors (Dudgeon et al., 2006; Ormerod & Durance, 2007; Ormerod et al., 2010). Moreover,
because the majority of riverine organisms are poikilothermic, they are affected metabolically
both by direct temperature change and by interactions between water temperature and oxygen
solubility (Graham & Harrod, 2009; Jonsson & Jonsson, 2009). Many freshwater ecosystems are
also coupled tightly to the surrounding riparian zones, floodplains and catchments through lateral
or longitudinal fluxes of energy that are under strong climatic influence (Nakano & Murakami,
2001; Wipfli, 2005). This includes the delivery, processing and downstream transport of detrital
carbon from terrestrial litter-fall that then acts as an important basis of production throughout
whole river systems (Vannote et al., 1980; Malmqvist, 2002).
Broad suggestions for adapting rivers to climate change are the same as for other ecosystems and
include enhancing resilience, connectivity and legal protection while reducing stressors such as
water quality impairment (Durance & Ormerod, 2009; Ormerod, 2009; Palmer et al., 2009).
However, some proposed adaptation strategies are specific to rivers such as reducing abstraction
(i.e. the active removal of water for human usage) and using riparian forest to buffer rivers
against temperature gain to protect sensitive organisms (Ormerod, 2009; Broadmeadow et al.,
Accepted Article
This article is protected by copyright. All rights reserved.
2011). Enhancing or restoring riparian tree cover is advocated particularly in temperate regions
where much native forest has been removed for agriculture (Battin et al., 2007; Palmer et al.,
2009; Seavy et al., 2009). Already, the value of such “buffer strips” in moderating stream
temperature is well supported by evidence (Zoellick, 2004; Battin et al., 2007; Broadmeadow et
al., 2011), some of it from our own study region (Weatherley & Ormerod 1990; Clews et al.
2010).
In addition to moderating thermal conditions in rivers, riparian woodlands might aid climate
change adaptation through effects on ecological processes, for example by soil denitrification,
nutrient flux and sediment delivery from agricultural land (Osbourne & Kovacic, 1993;
Broadmeadow & Nisbet, 2004; Larsen et al., 2009). More generally, riparian trees might affect
important aspects of stream and river energetics through two major pathways. First, shading
along streams is likely to reduce autotrophic productivity, potentially limiting resources for some
consumers (Hill et al., 1995; Kiffney et al., 2003; 2004; Riley et al., 2009). Secondly, increased
inputs of terrestrial organic matter from trees might provide important subsidies for consumers
linked to allochthony either in the form of abscised leaf litter (Wallace et al., 1997; Abelho, 2001)
or terrestrial invertebrates (Nakano & Murakami, 2001). Understanding any such collateral
effects on important river organisms such as macroinvertebrates could aid decisions on where and
when to use of riparian trees for shade and thermal damping. Additionally, allochthonous
energetic subsidies might increase stream ecosystem resilience to global change by increasing in-
stream biomass (Moore et al., 1993; Wallace et al., 1997; Muotka & Laasonen, 2002). There is a
need to assess whether smaller riparian ‘buffers’ of native woodland could provide such benefits
when used as climate change adaptation in the riparian zones of catchments managed for
agriculture or production forestry in an attempt to mimic more extensive woodland
(Broadmeadow & Nisbet, 2004; Wahl et al., 2013).
Accepted Article
This article is protected by copyright. All rights reserved.
A major difficulty in assessing the effects of riparian woodlands in climate change adaptation is
that several decades of tree growth are required between implementation of the concept and the
full realization of effects on stream systems. Elsewhere, however, we have used cross-sectional
comparison between sites with existing riparian broadleaves and other land uses to develop
predictions about possible effects on stream fishes (Thomas et al., 2015). The same study also
incorporated modern ecological methods – specifically stable isotopic analysis – as a means of
assessing energetic linkages between terrestrial subsidies and aquatic organisms (Rybczynski et
al., 2008; Ishikawa et al., 2012). Allochthonous and autochthonous production in streams is often
distinct enough isotopically to estimate their relative origins in freshwater organisms and hence to
appraise land use effects on their resource use (Doucett et al., 1996; Ishikawa et al., 2012). We
know of no study, however, where these or other techniques have been used to assess the
potential energetic effects of riparian adaptation strategies on macroinvertebrates – among the
most functionally important of all stream organisms. In combination, stable isotopic data,
quantitative estimates of macroinvertebrate biomass and taxonomic comparisons among streams
draining different land-use types can help to assess the putative consequences of variations in
riparian tree cover that could arise from climate change adaptation.
Here, we use cross-sectional comparisons at two scales (region and within-streams) among
replicate temperate streams in contrasting land use to test the hypothesis that climate change
adaptation using broadleaves can modify macroinvertebrate function and composition by
changing energetic pathways. Specific predictions were that (i) streams draining deciduous
woodland would be characterised by an increased abundance and biomass of leaf-shredding
invertebrates, due to increased inputs of terrestrial organic matter; (ii) resource use in
Accepted Article
This article is protected by copyright. All rights reserved.
invertebrates in deciduous woodland streams would reflect terrestrial production more than in
grassland streams; and (iii) riparian deciduous ‘buffers’ would approximate the effects on
invertebrate composition and resource use in more extensive catchment woodland.
Materials and Methods
Study sites
Sites were located in and around the Brecon Beacons National Park, South Wales, UK (51° 51'
46'' N, 3° 22' 41'' W Fig SM1) and the area has been described previously (Thomas et al., 2015).
Briefly, the region is temperate (1.1 °C - 19.1 °C mean min to mean max temperature; mean
annual rainfall is 1433 mm), with brown earth, gleys and occasionally peaty soils that mostly
overlay Devonian Old Red Sandstone drained by unpolluted, circumneutral and oligotrophic
headwaters (pH: ~6.5 – 7.5; conductivity: ~20 – 400 μS; Ca
2+
: ~5 – 40 mg l
-1
; NO
3
-: ~1 – 10 mg l
-
1
; PO
43-
: ~0 mg l
-1
). Temperate deciduous woodlands would form the climax vegetation, but most
land is now used for rough sheep grazing and commercial forestry with non-native conifers. As
such, the area is generally representative of upland land use patterns throughout the UK and
western Europe more generally. Moreover, such habitats represent ideal candidates for
management adaptations, as they are predicted to experience summer temperature increases of
around 4 – 5 °C by the 2080s (compared to historical averages) as a direct result of climate
change (UKCP09 medium emissions scenario; Murphy et al., 2009). Warming effects are already
apparent in the region, with increases in mean temperatures of 1.4 – 1.7 ˚C over the 25 years
between 1980 and 2005 (Durance and Ormerod, 2007), leading to local species extinction
(Durance & Ormerod 2010).
Accepted Article
This article is protected by copyright. All rights reserved.
Accepted Article
This article is protected by copyright. All rights reserved.
Series ICP-MS: Thermo Fisher Scientific, Inc.) and b) anions by ion chromatography (Dionex
DX-80 Ion Analyser; Thermo Fisher Scientific, Inc.). Conductivity, pH and total dissolved solids
(ppm) were assessed immediately following a storm event in October 2011, as these values are
typically at their most extreme during high flow (Kowalik et al., 2007). Three replicate readings
were taken at each 10m interval within each study reach using a Hanna HI 98129 low-range
pH/Conductivity/TDS Tester (Hanna Instruments, Ltd.)
Regional macroinvertebrate communities
During May-June 2010, benthic macroinvertebrates were collected from a 30 m reach at all 24
sites by separate kick-sample (D-frame kick net: net mesh 1 mm) respectively in riffles (i.e. fast-
flowing environments in the centre of the stream; 2 minute sampling duration) and marginal
habitats (i.e. slow-flowing depositional areas within 1m of the stream bank; 1 minute sampling
duration), and preserved in 70 % ethanol. This standardised procedure collects around 70 % of
species present at any one site and sufficient to detect differences among similar hillstreams, and
is based on methods developed over 25 years (Bradley & Ormerod, 2002). Separating riffle and
marginal samples provide a more representative species pool while also recording communities in
contrastingly eroding/depositing environments where Coarse Particulate Organic Matter (CPOM)
might accumulate (Bradley & Ormerod, 2002).
Kick-sample contents were sieved at 500 µm, sorted and identified as far as was practically
feasible, mostly species or genus except for Diptera (Athericidae, Ceratopogonidae,
Chironomidae, Pedicidae, Simuliidae, Tabanidae, Tipulidae) and some Coleoptera (Dytiscidae,
Accepted Article
This article is protected by copyright. All rights reserved.
Gyrindiae, Scirtidae), which were identified to family. Annelida were identified to subclass.
Ephemeropterans collected from marginal areas at site MO2 deteriorated during storage and this
site was excluded from some analyses. Using available data bases, taxa were assigned to one of
five functional feeding groups, according to the classification of Cummins and Klug (1979):
“Shredders” process coarse particulate organic matter (CPOM: principally decaying leaf litter and
riparian grasses); “Grazers” are primarily dependent on in-stream primary production,
predominantly epilithic algae; “Collector-Gatherers”, referred to as detritivores under some
classifications (Moog, 1995), utilise benthic fine particulate organic matter (FPOM); “Filterers”
obtain suspended materials from the water column; “Predators” capture and consume other
animal taxa (Moog 1995; Meritt and Cummins 1996; Hauer & Lamberti 2006).
Macroinvertebrate and CPOM within streams
In February, June and October of 2011 and 2012 (i.e. 6 occasions), macroinvertebrates and
CPOM standing stock were collected from fast-flowing riffles at the smaller sub-set of eight of
the 24 sites in 5 x 0.07 m
2
quantitative Hess samplers (Hess, 1941; upstream net: 1 mm mesh;
downstream net: 500 µm mesh; EFE-UK and GB Nets Ltd., UK). Samples were immediately
preserved in 70% Industrial Methylated Spirits (IMS: Fisher Scientific UK). Following treatment
as above for kick samples, all macroinvertebrate individuals from each taxon and sample were
transferred to glass vials for drying at 60 ˚C for 48 h and weighing to the nearest 0.1 mg.
Biomass data were expressed per m
2
of streambed. CPOM, defined as all non-woody vascular
plant material > 1 mm
2
(Cummins, 1974), was rinsed from each sample into a 1 mm sieve, and
also dried, weighed and converted to m
2
estimates of standing stock.
Accepted Article
This article is protected by copyright. All rights reserved.
Stable isotopes
Samples for stable isotope analysis were collected from all 24 study sites twice over the annual
cycle in May - June 2010 and again during January 2011. Benthic macroinvertebrates came from
kick-samples from which dominant macroinvertebrate taxa representing each major Functional
Feeding Group were removed on the bank-side, transferred to screw-top plastic vials and frozen
at -18 ºC within 8 hours. Later-instar individuals were collected preferentially to minimise effects
of ontogenetic dietary shifts (Dobson & Hildrew, 1992). Grazers were represented by heptageniid
and baetid mayflies; Shredders by leuctrid and nemourid stoneflies along with the amphipod
Gammarus pulex; Filterers by the Hydropsychidae (Trichoptera) and Simullidae (Diptera); and
Predators by the Perlidae, Chloroperlidae (Plecoptera) and Rhyacophilidae (Trichoptera).
Aggregate CPOM samples, mostly decaying broadleaf litter or riparian grasses from terrestrial
production, were collected simultaneously from the streambed, while epilithic biofilm (hereafter,
epilithon) representing in-stream primary production was scraped from the upper surfaces of
streambed rocks. Samples were frozen as above. Based on invertebrate body size (10 – 100 mg),
average stream water temperatures during the collection period (May/June: ~10 °C; January; ~5
°C) and turnover equations presented elsewhere (see Thomas & Crowther, 2015), the isotopic
composition of the selected consumers likely represented a relatively short-term integrator of
their seasonal resource use (estimated
13
C half-life: ~ 13 – 25 days). The chosen sampling
schedule should therefore have been sufficient to allow for detection of seasonal dietary shifts in
these taxa, if present.
All samples for stable isotope analysis were rinsed with DH
2
O and any non-target materials
removed using forceps before freeze-drying at -60 ºC for 48 h in glass vials. Dried samples were
Accepted Article
This article is protected by copyright. All rights reserved.
homogenized and quantities for analysis (1 ± 0.2 mg for invertebrate tissue, 3 ± 0.2 mg for
autotrophic material) were packaged within tin capsules for transport to the University of
California, Davis Stable Isotope Facility. Dual δ
13
C and δ
15
N analysis was performed using a
PDZ Europa ANCA-GSL elemental analyser interfaced to a PDZ Europa 20–20 isotope ratio
mass spectrometer (Sercon Ltd., Cheshire, U.K.). Values are reported in delta (δ) notation, as
parts per thousand (‰) deviation from international standards (Vienna Pee Dee Belemnite for
carbon and atmospheric air for nitrogen). Epilithic δ
15
N from DE4 was anomalously enriched (>
13‰ versus a mean of 1.42 ‰ at all other sites), probably reflecting local drainage, and was
excluded from analyses.
Statistical analysis
Statistical analyses were conducted in R Version 2.15.2 (R Development Core Team, 2012).
Initial analysis involved a combination of principal components and Analysis of Variance to
confirm expected land use variations among sites, and to appraise potential confounding
influences from other physico-chemical factors. These were mostly minor, although moorland
sites were at higher altitudes than others by ca. 150m while sites in Deciduous woodlands tended
to have higher conductivity reflecting generally increased ionic richness (Supplementary
material). These possible confounding effects are addressed below.
Regional macroinvertebrate communities
Accepted Article
This article is protected by copyright. All rights reserved.
Variations in community composition among land uses were initially plotted using Non-Metric
Multidimensional Scaling (NMDS; Kruskal, 1964) using the metaMDS function within R’s
vegan package (version 2.0-5), based on 500 iterations (Oksanen et al., 2012). NMDS is a robust
and well-known method that ordinates samples on overall dissimilarity (Kruskal, 1964; Clarke &
Warwick, 2001; Zuur et al., 2007), and was here used in conjunction with the Bray-Curtis index,
due to the ability of this metric to accommodate zero-skewed composition data (Clarke &
Warwick, 2001). All values were fourth root transformed prior to calculation to down-weight the
influence of the most abundant taxa (Clarke & Warwick, 2001). Permutational Multivariate
Analysis of Variance (PERMANOVA; Anderson, 2001) was subsequently used to assess whether
variations in community composition between land use types were significant. This non-
parametric alternative to MANOVA compares groups in multivariate space based on
dissimilarities and generates p values via a permutation procedure. PERMANOVA makes few
major assumptions about the data set, and does not require multivariate normality (Anderson,
2001). In order to rule out potentially confounding effects of differential dispersion,
PERMANOVAs were followed by betadisper tests (Anderson, 2006), a multivariate analogue of
Levene’s test for homogeneity of variances. Following an overall PERMANOVA to assess
whether land use affected community composition, we appraised group-by-group pair-wise
differences using the adonis function within vegan based on 4999 permutations (Oksanen et al.,
2012) following 4
th
root transformation. Where PERMANOVAs indicated significant differences
among land uses, Similarity Percentage analysis (SIMPER; Clarke, 1993) assessed which taxa
were principally responsible.
At the full suite of sites, we used General Linear Models (GLMs) to assess variations in total
macroinvertebrate abundance, diversity (Shannon) and FFG representation among land use
categories. Where PCAs indicated significant variations among land use categories differed in
Accepted Article
This article is protected by copyright. All rights reserved.
water chemistry or physical variables (see Supplementary Material), effects were controlled by
first modelling dependent variables against abiotic covariates (mean pH, mean conductivity, PC1
scores from anion and cation data, elevation, mean depth, mean width, catchment area, distance
from source), with stepwise deletion then used to remove all non-significant variables. Any
remaining significant terms for each dependent variable were included as covariates in each GLM
carried out to test for differences between land use categories.
Macroinvertebrate and CPOM within streams
At the eight sites sampled repeatedly, General linear mixed effects models (GLMMs; lme
function within the nlme package, Pinheiro et al., 2013) were used to assess differences in
macroinvertebrate biomass between land use types site-pairs and sampling periods, with site
fitted as a random term, in order to account for non-independence of samples taken from the same
location. Separate models were fitted to assess effects on total macroinvertebrate biomass, total
macroinvertebrate density, FFG-by-FFG biomass and proportional representation, and CPOM
standing stocks, with models including Land Use Type, Month and Year as explanatory variables,
along with all relevant two-way (including Month:Year, to investigate sampling-period-specific
differences), and three-way interactions. Where overall terms were significant, factor levels were
compared using Tukey’s Honestly Significant Difference (HSD) post-hoc comparisons.
The relationships between total macroinvertebrate biomass, total macroinvertebrate density, FFG-
by-FFG biomass, FFG-by-FFG proportional representation and the quantity of CPOM within
samples and were assessed using GLMMs. CPOM biomass was fitted as a covariate, along with
land use type, month and year as categorical explanatory variables, with all relevant interactions,
up to four-way, included. Site was again fitted as a random term to account for non-
Accepted Article
This article is protected by copyright. All rights reserved.
independence. Where necessary, variables for all models were log, log + 1, square root or Box-
Cox transformed prior to analysis, to meet linear model assumptions of normally distributed,
homoscedastic residuals and lack of autocorrelation. Because they were proportions, FFG
representation data were arcsine square root transformed (Sokal & Rohlf, 1995).
Stable isotopes
Our stable isotopic analysis depends on the assumption that the isotopic composition of
consumer tissues, particularly ratios of
13
C/
12
C and
15
N/
14
N, can indicate community-wide
dependence distinct food resources from different origins (Post, 2002; Layman et al., 2012).
When applied to different taxa within a food web, isotopic signatures are then used to infer
energy flow (Layman et al., 2012). In streams and rivers, this includes tracing back the energy
sources supporting macroinvertebrate consumers to their terrestrial (allochthonous) or in-stream
(autochthonous) origins, which are often isotopically distinct (Ishikawa et al., 2012).
Dual stable isotopic assessments of δ
13
C and δ
15
N was used here in conjunction with R’s SIAR
(Stable Isotope Analysis in R; version 4.1.3) mixing model (Parnell et al., 2010) with the
SIARsolomcmcv4 function to estimate proportional contribution from terrestrial and in-stream
production to consumer diets individually by site, functional feeding group and season. Mixing
models were fitted for 14 sites where basal resources were isotopically distinct (GB, n = 3; CB, n
= 4; DE, n = 3; MO, n = 4), but 10 sites were excluded where basal resource signatures
overlapped or where consumer measurements fell outside the mixing polygons implied by basal
signatures. All SIAR models were based on 500,000 iterations, with the first 50,000 discarded
Accepted Article
This article is protected by copyright. All rights reserved.
(Parnell et al., 2010). Trophic enrichment factors (TEFs) of 0.5 ± 0.5 ‰ for
13
C and 3.23 ± 1 ‰
for
15
N were assumed for primary consumers (Filterers, Grazers, Shredders) based on calculated
mean difference between primary consumers and basal resources. An additional trophic level of
enrichment was added for Predators (i.e. TEFs of 1 ± 1 ‰ and 6.46 ± 2 ‰ were used for
13
C and
15
N, respectively).
Variations in mean proportional contributions of terrestrial organic matter to consumer
production estimated by SIAR (hereafter, ‘terrestrial resource use’) were analysed using a
General Linear Mixed Model (GLMM). Riparian land use, month of sampling and Functional
Feeding Group, along with all possible interactions between these factors, were included as fixed
effects. Site was included as a random term to account for potential non-independence due to
repeated measures at each site through time. As analysis of all proportion data resulted in
normally distributed, homoscedastic residuals, no transformations were applied (Warton & Hui,
2011).
Results
Regional macroinvertebrate communities
Macroinvertebrate communities at the 24 regional sites varied among land uses in riffle (F
3,22
=
1.7442, p = 0.004), marginal (F
3,21
= 2.1634, p > 0.001) and combined samples (F
3,21
= 2.116, p >
0.001), with contrasts greatest between Moorland vs. Deciduous, Coniferous vs. Moorland and
Coniferous vs. Deciduous sites (Table 1; Fig 1). Buffer sites were generally intermediate,
Accepted Article
This article is protected by copyright. All rights reserved.
although communities in their marginal habitats differed from Deciduous sites (Table 1, 2).
SIMPER showed that differences in community composition were mostly due to overall changes
in abundance: no single taxon contributed > 7 % of the difference between any two land use
categories and those responsible represented a relatively small proportion of the total species pool
(Table 2). For example, differences between Deciduous sites and other land uses were principally
caused by increased abundance of the amphipod Gammarus pulex, decreases in the grazing
mayfly Electrogena lateralis and variations among Leuctra stoneflies. Conifererous sites differed
from others mostly because of increased abundances of leuctrid and nemourid stoneflies, notably
Amphinemura sulcicollis. Betadisper tests confirmed that differences between land uses were not
due to unequal dispersion between groups (Riffle: F
3, 19
= 0.140, p = 0.935; Margin: F
3, 18
= 0.326,
p = 0.807; Combined: F
3, 18
= 0.049, p = 0.985).
Functional group representation varied more strongly among land uses (Table 3; Table SM2), and
shredders contributed more to communities at Deciduous sites than any other land use in riffle,
margin and combined samples (Tukey’s HSD for Riffle: Deciduous v Buffer p = 0.008,
Coniferous p = 0.006, Moorland p = 0.005; Margin: Deciduous v Buffer p = 0.030, Coniferous p
= 0.005, Moorland p = 0.013; Combined: Deciduous v Buffer p = 0.011, Coniferous, p = 0.004,
Moorland p = 0.008). Other effects were weaker: Coniferous sites contained a higher proportion
of Grazers (p = 0.025) and Predators (p = 0.022) than at Deciduous sites, and a lower proportion
of Collector-Gatherers (p = 0.024) than at Moorland sites.
Macroinvertebrate and CPOM within streams
Accepted Article
This article is protected by copyright. All rights reserved.
At the 8 sites sampled repeatedly, benthic CPOM in Hess samples varied significantly among
riparian land uses (F
3, 213
= 43.41, p < 0.001) with amounts greater at Deciduous sites than in any
other site type (Tukey’s HSD: p < 0.001 in all cases), and lowest in Moorland (Fig. 2). Standing
stocks at Coniferous and Buffer sites were intermediate, and did not differ significantly from each
other (p = 0.557). These differences were maintained throughout the study, and did not depend
on sampling year (F
3, 213
= 1.04, p > 0.377) or month (F
6, 213
= 1.15, p > 0.337) despite some
seasonal variations (Fig 2).
Consistent with the variations in CPOM, total macroinvertebrate biomass (F
3, 213
= 14.57, p <
0.001) and density (F
3,213
= 15.84, p < 0.001) were both greater at Deciduous sites and lower at
Coniferous sites (Tukey’s HSD: p < 0.01 in all cases) than in any other land use (p < 0.05 in all
cases) when averaged across all sampling periods (Table 4; Fig 3). Moorland and Buffer sites had
intermediate biomass, and did not differ significantly from each other (p = 0.971). Again, these
effects occurred irrespective of variations in biomass and density among seasons and years (see
Supplementary Material Appendix SM2).
Biomass values in each FFG also varied among land uses when averaged across sampling periods
(Table 4). Shredder biomass was higher in Deciduous streams than in all other land uses (p < 0.05
in all cases), which did not differ significantly from one another (p > 0.05 in all cases). Collector-
Gatherers and Filterers had their lowest biomass in Coniferous streams (Tukey’s HSD: p < 0.05
in all cases), while Grazer biomass was also significantly lower in Coniferous than Deciduous
streams (p = 0.03). Predator biomass was higher in Moorland than Coniferous sites (p = 0.002),
but otherwise did not differ among land uses (p > 0.05 in all cases). Land use effects on FFG
biomass were consistent among months, years and individual sampling periods for Collector-
Accepted Article
This article is protected by copyright. All rights reserved.
Gatherers (Month: F
6, 213
= 0.90, p = 0.498; Year: F
3, 213
= 1.04, p = 0.377; Sampling Period: F
6, 213
= 1.26, p = 0.277), Predators (Month: F
6, 213
= 0.50, p = 0.810; Year: F
3, 213
= 2.54, p = 0.058;
Sampling Period: F
6, 213
= 2.24, p = 0.051) and Shredders (Month: F
6, 213
= 1.10, p = 0.365; Year:
F
3, 213
= 1.42, p = 0.238; Sampling Period: F
6, 213
= 1.26, p = 0.276). Variations among land uses
for Filterer (F
6, 213
= 2.18, p = 0.047) and Grazer biomass (F
6, 213
= 4.31, p < 0.001) were more
transient, and both differed among sampling months. The biomass of several functional feeding
groups also varied seasonally when averaged across land use types (see Supplementary material
Appendix SM2). In general, similar patterns were confirmed by proportionate variations among
FFGs, and in particular Deciduous sites had a greater proportion of Shredder taxa than all other
land use types (p < 0.001 in all cases) while Coniferous sites supported a lower proportion of
Collector-Gatherer taxa and greater proportion of Grazers (Tukey’s HSD: p < 0.05 in all cases)
(See Appendix SM3).
Supporting a likely effect of CPOM on macroinvertebrates across land uses, total
macroinvertebrate biomass (F
1, 189
= 94.96, p < 0.001) and density (F
1, 189
= 138.63, p < 0.001)
both increased significantly in samples with greater standing stocks of CPOM (Fig. 4; Table 5).
These relationships were independent of land use type (Biomass: F
3, 189
= 2.49, p = 0.062;
Density: F
3, 189
= 0.53, p = 0.661), month (Biomass: F
2, 189
= 0.41, p = 0.665; Density: F
2, 189
= 2.12,
p = 0.122) or year (Biomass: F
1, 189
= 0.74, p = 0.393; Density: F
1, 189
= 0.02, p = 0.888). Within
individual guilds, Shredder biomass also increased with CPOM biomass across samples (F1, 189
= 7.63, p = 0.006), though the relationship varied seasonally (F
2, 189
= 5.85, p = 0.003). Similarly,
the proportion of total macroinvertebrate biomass composed of Shredders was significantly
positively related to CPOM biomass (F
1, 189
= 17.22, P < 0.001), but the relationship varied
between months (F
1, 189
= 4.52, p = 0.012) and years of sampling (F
1, 189
= 9.93, p = 0.002). The
Accepted Article
This article is protected by copyright. All rights reserved.
biomass or proportional representation of all other functional feeding groups was not significantly
related to CPOM biomass.
Stable isotopes
Contrary to prediction (ii) and unexpectedly given the apparent relationship between land use,
CPOM and macroinvertebrate biomass, terrestrial resource use by macroinvertebrates, as revealed
by isotopic data, did not vary significantly among riparian land use types (F
3, 95
= 0.416, p =
0.742; Fig. 5) even when variations between months (F
3, 93
= 0.923, p = 0.433) or FFGs (F
8, 87
=
0.620, p = 0.759) were considered. Across all land use categories in both months, roughly 50 %
(range: 33.1 - 75.8 %) of resources assimilated by all macroinvertebrate functional groups were of
terrestrial origin (Fig. 5). When all land use categories were pooled, terrestrial resource use
varied between functional feeding groups in ways that differed between months (F
3, 95
= 3.890, p
= 0.012), but this effect occurred only as significantly increased terrestrial contributions to Grazer
tissues in June (p = 0.002; Fig SM 3).
Discussion
Despite increasing concern about climate change, practical evidence about the effectiveness of
management adaptations that could reduce adverse effects on ecosystems is still remarkably
scarce. To our knowledge this study, combined with an associated paper (Thomas et al., 2015), is
the first to appraise collateral ecological effects of using riparian trees to create shade - one of the
most widely advocated adaptation measures for rivers (Ormerod 2009; Palmer et al., 2009; Clews
Accepted Article
This article is protected by copyright. All rights reserved.
et al., 2010). Of the three predictions we examined, only one was supported unequivocally:
streams draining deciduous woodland differed clearly from others in having substantially
enhanced standing stocks of CPOM as well as a greater density and biomass of
macroinvertebrates, particularly Shredders. In contrast, there were no variations across land uses
in functional group reliance on terrestrial resources (prediction ii), and nor were the effects of
riparian buffers of 15-60m width sufficient to mimic the effects of more extensive riparian
woodlands (prediction iii). These outcomes provide some support for the hypothesis that climate
change adaptation using broadleaves might alter macroinvertebrate communities functionally and
compositionally, also enhancing stocks of CPOM as an important basal resource. However, on
our evidence this effect is likely only where broadleaf restoration or planting is extensive, and
large step-changes from autochthony to allochthony may not be a major feature. Interestingly,
the data support previous suggestions that narrow riparian buffer zones may be insufficient to
offset some of the influences of wider catchment land use on stream communities and ecosystem
functioning (Allan et al., 1997; Kauffman et al., 1997; Harding et al., 2006; Wahl et al., 2013).
Before discussing more general ramifications of this study, two important caveats must be noted.
First, as with other cross-sectional investigations using space-for-time substitution, our site
categories were created neither by experimental manipulation nor random allocation to
treatments. Our interpretation must, therefore, rely on correlative techniques that are at risk from
possible confounding effects. The land use categories differed marginally on physicochemistry,
with, for example, treeless moorland streams at higher elevations than other land use types
(Supplementary material). However, the range over which these variables differed appeared to be
insufficient to influence community composition: moorland (MO) and buffer strip (GB) sites
differed on physical criteria but supported similar communities. Similarly, buffer strip (GB) sites
and those draining larger areas of deciduous woodland (DE) differed with respect to water
Accepted Article
This article is protected by copyright. All rights reserved.
chemistry, but not in overall macroinvertebrate community composition and the principal results
were consistent across the two scales of the investigation. In other fields, such as freshwater
acidification, early evidence based on space-for-time substitution (Ormerod et al. 1988) has since
been validated using long-term data (Ormerod & Durance 2009) and coupled with studies of
processes (Kowalik et al. 2007) to provide important insights into global change effects. A second
caveat is that the study was intended to appraise the effects of riparian management as an
adaptation to climate change, yet could only be carried out under current climatic conditions.
While there is already evidence of warming effects on streams in the study region (Durance &
Ormerod 2007, 2010; Clews et al., 2010), any extrapolation requires the assumption that patterns
detected here will persist under the higher temperatures, more variable rainfall and potentially
extreme discharge expected in NW Europe. Interestingly, future climates could also affect
streamside woodlands as well as in-stream conditions – for example through altered disease
effects on tree species such as European Ash (Fraxinus excelsior) (Pautasso et al., 2013).
Notwithstanding these concerns, we suggest that comparative studies like ours provide a useful
basis for predicting how temperate upland streams might respond to the restoration of catchment
tree cover, thereby increasing understanding of the resultant ecological changes (Naiman et al.,
2012). Decades would be required for the experimental development of riparian tree cover, yet
evidence to inform decision about climate change adaptation are required now.
The clearest overall trends we detected, principally increased CPOM stocks, enhanced shredder
density and increased macroinvertebrate biomass in extensive deciduous woodlands, are
generally well known (Wallace et al., 1997). Interestingly, these effects occurred in both riffles
and margins, but were stronger in the latter, where leaf litter and other terrestrial organic material
often accumulate in ‘softer’ habitats of woody debris, roots and vegetated features (Ormerod et
al., 1993; Flores et al., 2013). Such marginal features that increase riparian shore-length and
Accepted Article
This article is protected by copyright. All rights reserved.
increase litter retention could be as important in increasing CPOM stocks as the adjacent canopy
is in providing inputs (Muotka & Laasonen, 2002). Deciduous woodlands would form the
principal climax vegetation communities over large parts of the temperate zone and, where
riparian zones are intact, the resulting litter input to headwaters is a key component of energy flux
through food webs (Vannote et al., 1980). Where subsidies of CPOM are large enough to offset
climatically mediated export, riparian woodlands might also increase the resilience of
macroinvertebrate populations by increasing basal resources (Moore et al., 1993; Wallace et al.,
1995; Eggert et al., 2012). Understanding future interactions between litter subsidies, uptake into
food webs, secondary production and climate change is likely to be a key area of interest: land
use in the riparian zone is important in that it mediates both climatic effects and riparian subsidies
(Wallace et al., 1995; Broadmeadow et al., 2011).
Despite apparently changing CPOM and invertebrate abundance across the study streams,
however, land use did not affect the relative use of terrestrial and aquatic resource use by
macroinvertebrates in any functional feeding group. Thus, while shredders apparently intercepted
the terrestrial subsidy at deciduous sites and converted it into increased invertebrate biomass, they
still depended in part on autotrophic production. This contrasts with the resource-use patterns
typically assigned to this group (Cummins & Klug, 1979). Such effects would arise where
shredding taxa ingested and assimilating algal production attached to leaf litter (Hax & Golladay,
1993). In the same way, grazing taxa can supplement their diets with fine terrestrial organic
matter captured within epilithic biofilms (Hamilton et al., 2005). Thus, even at open moorland
sites where greater autotrophic production would be expected, roughly 50% of animal production
in all FFGs originated from terrestrial sources. These results are contrary to the expectation that
relative allocthony versus autochthony should differ among deciduous woodland, grassland and
conifer sites (Abehlo, 2001; Kiffney et al., 2003, 2004).
Accepted Article
This article is protected by copyright. All rights reserved.
Such unexpected effects might be explained either by intrinsic biological processes or by
extrinsic, contextual factors alone or combination. For example, autotrophic food webs persist
even in streams in woodland environments prior to seasonal canopy closure or where primary
production is maintained in habitats such as mosses (Wallace et al., 1997). Equally,
allochthonous resources still occur in moorland catchments, and Menninger & Palmer (2007)
illustrated how herbs and grasses provided significant inputs of litter to open-canopy piedmont
streams. Leberfinger et al. (2011) used stable isotope analysis to show that such terrestrial
organic resources were important to shredding macroinvertebrates in grassland streams despite
the availability of autotrophic production. Resource-use patterns might also be mediated directly
and locally by invertebrate consumers: despite varying amounts of terrestrial organic matter in
different land uses, use and uptake can be constrained by the capacity for feeding plasticity in
consumer taxa (e.g. morphological adaptations for rock scraping vs. leaf shredding vs. filter
feeding) (Cummins & Klug, 1979; Dangles, 2002). Variations in resource quality between
terrestrial and in-stream production might also constrain feeding choices, with CPOM typically
less rich in macronutrients than benthic epilithon: macroinvertebrates often require elemental
stochiometry with their food sources, and CPOM alone may be insufficient to support growth and
metabolism (Hladyz et al., 2009). More extrinsically, wider catchment effects or downstream
export might mask local riparian effects from land use: there is evidence to indicate that even
small reductions in catchment tree cover (~10% deforestation in otherwise totally afforested
catchments) weakens terrestrial-aquatic linkages (England & Rosemond, 2004). Resource use
patterns may therefore reflect whole catchment land use rather than those just in the riparian zone,
even where wider riparian land uses are extensive: in woodland catchments, large areas of lateral
tree cover may be needed to offset downstream subsidy export, particularly transport during high
flow events (Wallace et al. 1995; Eggert et al., 2012). Finally, measurement or modelling
Accepted Article
This article is protected by copyright. All rights reserved.
artefacts with stable isotopes cannot be excluded. For instance, in some riverine systems it may
be difficult to fully discern in-stream production using isotopic signals alone, as there can be
substantial overlap in
13
C concentrations between autochthnous and allochthonous material
(France, 1996). However, care was taken to exclude data from sites where assumptions of
isotopic mixing models were violated, and repeated sampling at each site improved robustness by
minimizing the effects of temporal variability. Additionally, potential error in the estimation
resource isotopic composition was of explicitly incorporated into the Bayesian models used here
(Parnell et al., 2010).
Implications for climate change adaptation
A central theme of our study is an appraisal of how the protection, management and restoration of
riparian broadleaves for climate-change adaptation might have effects on temperate headwaters
beyond cooling alone. Already, there is extensive investment in Britain and elsewhere to instigate
riparian tree planting based on evidence that the resulting shade dampens thermal variation in
adjacent streams (Broadmeadow & Nisbet, 2004; Broadmeadow et al., 2011; Environment
Agency, 2011). Here, we set out to assess whether there might also be collateral effects on
important basal resources or aspects of stream function mediated by macroinvertebrates. Our
previous paper using data from some of the same locations showed that deciduous riparian zones
were neutral for salmonids, but conifers reduced density and biomass (Thomas et al., 2015). The
latter result is consistent with data collected here in that overall macroinvertebrate biomass and
density were also lowest in coniferous catchments.
Accepted Article
This article is protected by copyright. All rights reserved.
Small-scale, riparian interventions in catchments used otherwise for agriculture or urbanization
are postulated often as a potentially valuable and cost-effective means of reducing warming
directly, reducing stressors that could be exacerbated by warming (e.g. nutrients, sediments) and
enhancing resilience by providing habitat and energetic subsidies (Moore et al., 1993; Wallace et
al., 1997; Sweeney et al., 2004). Our data extend understanding this technique by illustrating how
riparian broadleaves used in climate change adaptation would be likely to enhance CPOM stocks,
Shredder densities and overall macroinvertebrate biomass, but only where woodland is extensive.
In this respect the management implications are clear. First, where narrow woodland buffers are
used solely to moderate thermal regimes, large energetic benefits or effects on CPOM dynamics
would, on our evidence, be unlikely. In contrast, more extensive woodland restoration would
have additional further benefits beyond cooling through enhanced litter subsidies and retention in
marginal habitats. Both of these effects could be enhanced by encouraging planting or natural
regeneration of native riparian vegetation, for example through agri-environment schemes or as
part of wider global reforestation efforts (Crowther et al., 2015), and by protecting retentive
features such as marginal vegetation or woody debris during river management (Muotka &
Laasonen, 2002; Sweeney et al., 2004; Flores et al. 2013). Such techniques could either be
implemented alone, or in combination with other adaptation strategies which could act to enhance
local shade and cooling (e.g. Everall et al., 2012). The ecological benefits locally within
headwaters might include improved conservation of woodland stream organisms, restored linkage
between headwaters and riparian zones, restoration of natural stream function and potentially
increased resilience through enhanced basal resources (Moore et al., 1993; Goodwin et al., 1997;
Muotka & Laasonen, 2002). More extensive benefits from litter processing and export to
ecosystems downstream are possible, but require fuller appraisal in more extensively wooded
landscapes (Wipfli, 2005; Tanentzap et al., 2014).
Accepted Article
This article is protected by copyright. All rights reserved.
More generally, our study illustrates a possible approach to appraising the potential effectiveness
of climate change adaptation – through studies of ecological processes combined with a survey of
locations assumed to mimic future land cover options. With current practical knowledge of
climate change adaptation in most ecosystems still rudimentary, we suggest that this subject
needs greater research attention given the urgency for action and the time required to develop,
implement and fully realize change at the landscape scales necessary.
Acknowledgements
We thank the Knowledge Economy Skills Scholarship scheme for funding SMT, and the NERC
‘Duress’ and EU MARS projects for funding SJO. We also thank the South East Wales Rivers
Trust and Natural Resources Wales for help in kind, and Caitlin Pearson, Matthew Dray and
Stuart Rudd for assistance in the field. Three anonymous referees provided insightful comments
on the manuscript.
References
Abelho M (2001). From litterfall to breakdown in streams: a review. The Scientific World
Journal, 1, 656-680.
Allan D, Erickson D & Fay J (1997). The influence of catchment land use on stream integrity
across multiple spatial scales. Freshwater Biology, 37, 149-161.
Accepted Article
This article is protected by copyright. All rights reserved.
Anderson MJ (2001). A new method for non-parametric multivariate analysis of variance. Austral
Ecology, 26, 32–46.
Anderson MJ (2006) Distance-based tests for homogeneity of multivariate dispersions.
Biometrics, 62, 245–253.
Battin J, Wiley MW, Ruckelshaus MH, Palmer RN, Korb E, Bartz KK & Imaki H (2007).
Projected impacts of climate change on salmon habitat restoration. Proceedings of the National
Academy of Sciences, 104, 6720-6725.
Bradley DC & Ormerod SJ (2002). Evaluating the precision of kick-sampling in upland streams
for assessments of long-term change: the effects of sampling effort, habitat and rarity. Archiv für
Hydrobiologie, 155, 199-221.
Broadmeadow S & Nisbet TR (2004). The effects of riparian forest management on the
freshwater environment: a literature review of best management practice. Hydrology and Earth
System Sciences Discussions, 8, 286-305.
Accepted Article
This article is protected by copyright. All rights reserved.
Broadmeadow SB, Jones JG, Langford TEL, Shaw PJ & Nisbet TR. (2011). The influence of
riparian shade on lowland stream water temperatures in southern England and their viability for
brown trout. River Research and Applications, 27, 226-237.
Clarke KR (1993). Non‐parametric multivariate analyses of changes in community structure.
Australian Journal of Ecology, 18, 117-143.
Clarke KR & Warwick RM (2001). Changes in Marine Communities: an Approach to Statistical
Analysis and Interpretation. Second edition. PRIMER-E, Plymouth, UK.
Clews E, Vaughan IP & Ormerod SJ (2010). Evaluating the effects of riparian restoration on a
temperate river‐system using standardized habitat survey. Aquatic Conservation: Marine and
Freshwater Ecosystems, 20(S1), S96-S104.
Countryside Council for Wales (2004). Phase 1 habitat survey.
Crowther TW, Glick HB, Covey KR, et al. (2015). Mapping tree density at a global scale.
Nature, doi:10.1038/nature14967
Cummins KW & Klug MJ (1979). Feeding ecology of stream invertebrates. Annual Review of
Ecology and Systematics, 10, 147-172.
Accepted Article
This article is protected by copyright. All rights reserved.
Cummins KW, Wilzbach MA, Gates DM, Perry JB & Taliaferro WB (1989). Shredders and
riparian vegetation. BioScience, 39, 24-30.
Cummins KW (1974). Structure and function of stream ecosystems. BioScience, 24, 631-641.
Dangles O (2002). Functional plasticity of benthic macroinvertebrates: implications for trophic
dynamics in acid streams. Canadian Journal of Fisheries and Aquatic Sciences, 59, 1563-1573.
Dobson M & Hildrew AG (1992). A test of resource limitation among shredding detritivores in
low order streams in southern England. Journal of Animal Ecology, 61, 69-77.
Doucett RR, Barton DR, Guiguer KR, Power G & Drimmie RJ (1996). Comment: Critical
examination of stable isotope analysis as a means for tracing carbon pathways in stream
ecosystems. Canadian Journal of Fisheries and Aquatic Sciences, 53, 1913-1915.
Dudgeon D, Arthington AH, Gessner MO et al. 2006. Freshwater biodiversity: importance,
threats, status and conservation challenges. Biological Reviews, 81, 163–182.
Durance I & Ormerod SJ (2007). Climate change effects on upland stream macroinvertebrates
over a 25‐year period. Global Change Biology, 13, 942-957.
Accepted Article
This article is protected by copyright. All rights reserved.
Durance I & Ormerod SJ (2009). Trends in water quality and discharge confound long-term
warming effects on river macroinvertebrates. Freshwater Biology, 54, 388-405.
Durance I & Ormerod SJ (2010). Evidence for the role of climate in the local extinction of a cool-
water triclad. Journal of the North American Benthological Society, 29, 1367-1378.
Eggert SL, Wallace JB, Meyer JL & Webster J. R. (2012) Storage and export of organic matter in
a headwater stream: responses to long-term detrital manipulations. Ecosphere, 3, Article UNSP
75.
England LE & Rosemond AD (2004). Small reductions in forest cover weaken terrestrial‐aquatic
linkages in headwater streams. Freshwater Biology, 49, 721-734.
Environment Agency (2011) Keeping Rivers Cool: Getting ready for climate change by creating
riparian shade. Environment Agency, Bristol .
Everall NC, Farmer A, Heath A F, Jacklin TE & Wilby RL (2012). Ecological benefits of
creating messy rivers. Area, 44, 470-478.
Accepted Article
This article is protected by copyright. All rights reserved.
Flores L, Díez JR, Larrañaga A, Pascoal C & Elosegi A (2013). Effects of retention site on
breakdown of organic matter in a mountain stream. Freshwater Biology, 58, 1267 -1278
France RL (1996). Carbon-13 conundrums: limitations and cautions in the use of stable isotope
analysis in stream ecotonal research. Canadian Journal of Fisheries and Aquatic Sciences, 53,
1916-1919.
Graham CT & Harrod C (2009). Implications of climate change for the fishes of the British Isles.
Journal of Fish Biology, 74, 1143-1205.
Goodwin CN, Hawkins CP & Kershner JL (1997). Riparian restoration in the western United
States: overview and perspective. Restoration Ecology, 5, 4-14.
Hamilton SK, Sippel SJ & Bunn SE (2005). Separation of algae from detritus for stable isotope or
ecological stoichiometry studies using density fractionation in colloidal silica. Limnology and
Oceanography: Methods, 3, 149-157.
Harding JS, Claassen K & Evers N (2006). Can forest fragments reset physical and water quality
conditions in agricultural catchments and act as refugia for forest stream invertebrates?
Hydrobiologia, 568, 391-402.
Accepted Article
This article is protected by copyright. All rights reserved.
Hauer FR & Lamberti GA (2006). Methods in Stream Ecology. Academic Press.
Hax CL & Golladay SW (1993). Macroinvertebrate colonization and biofilm development on
leaves and wood in a boreal river. Freshwater Biology, 29, 79-87.
Hess AD (1941). New limnological sampling equipment. Limnological Society of America.
Special Publication 6, 1 – 5.
Hladyz S, Gessner MO, Giller PS, Pozo J & Woodward G. (2009). Resource quality and
stoichiometric constraints on stream ecosystem functioning. Freshwater Biology, 54, 957-970.
Hill WR, Ryon MG & Schilling EM (1995). Light limitation in a stream ecosystem: responses by
primary producers and consumers. Ecology, 76, 1297-1309.
Holmlund CM & Hammer M (1999). Ecosystem services generated by fish populations.
Ecological Economics, 29, 253-268.
Hulme PE (2005). Adapting to climate change: is there scope for ecological management in the
face of a global threat? Journal of Applied Ecology, 42, 784-794.
Accepted Article
This article is protected by copyright. All rights reserved.
IPCC, 2014: Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global and
Sectoral Aspects. Contribution of Working Group II to the Fifth Assessment Report of the
Intergovernmental Panel on Climate Change [Field CB, Barros VR, Dokken DJ et al. (eds.)].
Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1132 pp.
Ishikawa NF, Doi H & Finlay JC (2012). Global meta-analysis for controlling factors on carbon
stable isotope ratios of lotic periphyton. Oecologia, 170, 541-549.
Jonsson B & Jonsson N (2009). A review of the likely effects of climate change on anadromous
Atlantic salmon Salmo salar and brown trout Salmo trutta, with particular reference to water
temperature and flow. Journal of Fish Biology, 75, 2381-2447.
Kauffman JB, Beschta RL, Otting N & Lytjen D (1997). An ecological perspective of riparian
and stream restoration in the western United States. Fisheries, 22, 12-24.
Kiffney PM, Richardson JS & Bull JP (2003). Responses of periphyton and insects to
experimental manipulation of riparian buffer width along forest streams. Journal of Applied
Ecology, 40, 1060-1076.
Kiffney PM, Richardson JS & Bull JP (2004). Establishing light as a causal mechanism
structuring stream communities in response to experimental manipulation of riparian buffer
width. Journal of the North American Benthological Society, 23, 542-555.
Accepted Article
This article is protected by copyright. All rights reserved.
Kowalik RA, Cooper DM, Evans CD & Ormerod SJ (2007). Acidic episodes retard the biological
recovery of upland British streams from chronic acidification. Global Change Biology, 13, 2439–
2452.
Kruskal JB (1964). Nonmetric multidimensional scaling: a numerical method. Psychometrika, 29,
115-129.
Larsen S, Vaughan IP & Ormerod SJ (2009). Scale–dependent effects of fine sediments on
temperate headwater invertebrates. Freshwater Biology, 54, 203-219.
Layman CA, Araujo MS, Boucek R et al. (2012). Applying stable isotopes to examine food‐web
structure: an overview of analytical tools. Biological Reviews, 87, 545-562.
Leberfinger K, Bohman I & Herrmann J (2011). The importance of terrestrial resource subsidies
for shredders in open‐canopy streams revealed by stable isotope analysis. Freshwater Biology, 56,
470-480.
Macgregor, N. & van Dijk, AN (2014) Adaptation in Practice: How Managers of Nature
Conservation Areas in Eastern England are Responding to Climate Change. Environmental
Management, 54, 700-719.
Accepted Article
This article is protected by copyright. All rights reserved.
Malmqvist B (2002). Aquatic invertebrates in riverine landscapes. Freshwater Biology, 47, 679-
694.
Mawdsley JR, O'Malley R & Ojima DS (2009) A Review of Climate-Change Adaptation
Strategies for Wildlife Management and Biodiversity Conservation. Conservation Biology, 23,
1080-1089
Menninger HL & Palmer MA (2007) Herbs and grasses as an allochthonous resource in open-
canopy headwater streams. Freshwater Biology, 52, 1689-1699
Merritt RW & Cummins KW (Eds.) (1996). An Introduction to the Aquatic Insects of North
America. Kendall Hunt.
Moog O (Ed.) (1995). Fauna Aquatica Austriaca. Wasser-Wirtschafts-Kataster,
Bundesministerium für Land-und Forstwirtschaft.
Moore JC, Deruiter PC & Hunt HW (1993) Influence of productivity on the stability of real and
model ecosystems. Science, 261, 906-908.
Muotka T & Laasonen P (2002) Ecosystem recovery in restored headwater streams: the role of
enhanced leaf retention. Journal of Applied Ecology, 39, 145-156
Accepted Article
This article is protected by copyright. All rights reserved.
Murphy JM, Sexton DMH, Jenkins GJ, et al. 2009 UK Climate Projections Science Report:
Climate Change Projections. Exeter, UK, Meteorological Office Hadley Centre, 192pp.
Naiman RJ, Alldredge JR, Beauchamp DA et al. (2012). Developing a broader scientific
foundation for river restoration: Columbia River food webs. Proceedings of the National
Academy of Sciences, 109, 21201-21207.
Nakano S & Murakami M (2001). Reciprocal subsidies: dynamic interdependence between
terrestrial and aquatic food webs. Proceedings of the National Academy of Sciences, 98, 166-170.
Oksanen J, Blanchet FG, Kindt R et al. (2012) vegan: Community Ecology Package. R package
version, 2.0-5.
Ormerod SJ (2009). Climate change, river conservation and the adaptation challenge. Aquatic
Conservation: Marine and Freshwater Ecosystems, 19, 609-613.
Ormerod SJ, Dobson M, Hildrew AG & Townsend CR (2010). Multiple stressors in freshwater
ecosystems. Freshwater Biology, 55(s1), 1-4.
Accepted Article
This article is protected by copyright. All rights reserved.
Ormerod SJ, Rundle SD, Lloyd EC & Douglas AA (1993). The influence of riparian management
on the habitat structure and macroinvertebrate communities of upland streams draining plantation
forests. Journal of Applied Ecology, 30, 13-24.
Ormerod SJ, Weatherley NS, Varallo PV & Whitehead PG (1988). Preliminary empirical models
of the historical and future impact of acidification on the ecology of Welsh streams. Freshwater
biology. Freshwater Biology, 20, 127-140.
Osborne LL & Kovacic DA (1993). Riparian vegetated buffer strips in water‐quality restoration
and stream management. Freshwater biology, 29, 243-258.
Palmer MA, Lettenmaier DP, Poff NL, Postel SL, Richter B & Warner R (2009). Climate change
and river ecosystems: protection and adaptation options. Environmental Management, 44, 1053-
1068.
Parnell AC, Inger R, Bearhop S & Jackson AL (2010). Source partitioning using stable isotopes:
coping with too much variation. PLOS ONE, 5, e9672.
Pautasso M, Aas G, Queloz V & Holdenrieder O (2013). European ash (Fraxinus excelsior)
dieback–a conservation biology challenge. Biological Conservation, 158, 37-49.
Accepted Article
This article is protected by copyright. All rights reserved.
Perry J (2015) Climate change adaptation in the world's best places: A wicked problem in need of
immediate attention. Landscape and Urban Planning, 133, 1-11.
Pinheiro J, Bates D, DebRoy S & Sarkar D (2013). R Development Core Team (2012) nlme:
linear and nonlinear mixed effects models. R package version 3.1-103. R Foundation for
Statistical Computing, Vienna.
Post DM (2002). Using stable isotopes to estimate trophic position: models, methods, and
assumptions. Ecology, 83, 703-718.
R Core Team (2012). R: A language and environment for statistical computing. R Foundation for
Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL: http://www.R-project.org/.
Riley WD, Pawson MG, Quayle V & Ives MJ (2009). The effects of stream canopy management
on macroinvertebrate communities and juvenile salmonid production in a chalk stream. Fisheries
Management and Ecology, 16, 100-111.
Rybczynski SM, Walters DM, Fritz KM & Johnson BR (2008). Comparing trophic position of
stream fishes using stable isotope and gut contents analyses. Ecology of Freshwater Fish, 17, 199-
206.
Accepted Article
This article is protected by copyright. All rights reserved.
Seavy NE, Gardali T, Golet GH et al. (2009). Why climate change makes riparian restoration
more important than ever: recommendations for practice and research. Ecological Restoration, 27,
330-338.
Sokal RR & Rohlf FJ (1995). Biometry (3rd edn). WH Freman and company: New York.
Sweeney BW, Bott TL, Jackson JK et al. (2004). Riparian deforestation, stream narrowing, and
loss of stream ecosystem services. Proceedings of the National Academy of Sciences, 101,
14132-14137.
Tanentzap AJ, Szkokan-Emilson EJ, Kielstra BW, Arts MT, Yan ND & Gunn JM (2014) Forests
fuel fish growth in freshwater deltas. Nature Communications, 5, Article Number: 4077.
Thomas SM & Crowther TW (2015). Predicting rates of isotopic turnover across the animal
kingdom: a synthesis of existing data. Journal of Animal Ecology, 84, 861–870.
Thomas SM, Griffiths SW & Ormerod SJ (2015). Adapting streams for climate change using
riparian broadleaf trees and its consequences for stream salmonids. Freshwater Biology, 60, 64–
77.
Vannote RL, Minshall GW, Cummins KW, Sedell JR & Cushing CE (1980). The river
continuum concept. Canadian Journal of Fisheries and Aquatic Sciences, 37, 130-137.
Accepted Article
This article is protected by copyright. All rights reserved.
Wahl CM, Neils A & Hooper D (2013). Impacts of land use at the catchment scale constrain the
habitat benefits of stream riparian buffers. Freshwater Biology, 58, 2310–2324.
Wallace JB, Eggert SL, Meyer JL & Webster JR (1997). Multiple trophic levels of a forest stream
linked to terrestrial litter inputs. Science, 277, 102-104.
Wallace JB, Whiles MR, Eggert S, Cuffney TF, Lugthart GJ & Chung K (1995). Long-term
dynamics of coarse particulate organic matter in three Appalachian Mountain streams. Journal of
the North American Benthological Society, 14, 217-232.
Warton DI & Hui FK (2011). The arcsine is asinine: the analysis of proportions in ecology.
Ecology, 92, 3-10.
Weatherley NS & Ormerod SJ (1990). Forests and the temperature of upland streams: a
modelling study of the biological consequences. Freshwater Biology, 24, 109-122.
Wilson MA & Carpenter SR (1999). Economic valuation of freshwater ecosystem services in the
United States: 1971-1997. Ecological Applications, 9, 772-783.
Wipfli MS (2005). Trophic linkages between headwater forests and downstream fish habitats:
implications for forest and fish management. Landscape and Urban Planning, 72, 205-213.
Accepted Article
This article is protected by copyright. All rights reserved.
Zoellick BW (2004). Density and biomass of redband trout relative to stream shading and
temperature in southwestern Idaho. Western North American Naturalist, 64, 18–26.
Zuur AF, Ieno EN & Smith GM (2007). Analysing Ecological Data. Springer.
Accepted Article
This article is protected by copyright. All rights reserved.
Accepted Article
This article is protected by copyright. All rights reserved.
Accepted Article
This article is protected by copyright. All rights reserved.
Accepted Article
This article is protected by copyright. All rights reserved.
Figure Legends
Figure 1: NMDS ordinations of macroinvertebrate communities (after 4th root transformation)
collected from South Wales streams in a.) riffles; b.) marginal habitats; c.) combined samples:
points indicate Buffer (solid lines; ), Coniferous (dashed lines; ), Deciduous (dotted lines; )
and Moorland (dot-dash lines; ) sites.
Figure 2: CPOM biomass (mg m
-2
: mean ± 1S.E.) dynamics across land use types and sampling
periods. a.) site-specific values averaged across all sampling periods, b.) yearly values averaged
across all sites, c.) monthly values averaged across sites in 2011 and d.) monthly values averaged
across sites in 2012.Land use categories: CB = Coniferous, DE = Deciduous, GB = Buffer, MO
= Moorland. Y-axis scales differ between graphs.
Figure 3: Macroinvertebrate biomass (mg m
-2
: mean ± 1 SE) over two years (2011 and 2012) at
eight streams in South Wales draining different land use: CB = Coniferous, DE = Deciduous, GB
= Buffer, MO = Moorland. Shared letters denote land use type site-pairs that did not differ
significantly within each sampling period (Tukey’s post-hoc comparisons following GLMM: p >
0.05).
Figure 4: Relationships between log transformed CPOM biomass and a.) total macroinvertebrate
biomass, b.) total macroinvertebrate density. Solid lines indicate best fit as predicted by Linear
Mixed Effects models, dashed lines represent predicted standard errors around the mean.
Accepted Article
This article is protected by copyright. All rights reserved.
Figure 5: Estimated proportional terrestrial resource use in each of four macroinvertebrate
functional groups collected for stable isotope analysis in streams in South Wales, across land use
types on two sampling occasions: a.) filtering taxa, b.) grazing taxa, c.) predatory taxa and d.)
shredding taxa. Values presented are mean proportional terrestrial resource use ± 1 SE derived
from SIAR.
Accepted Article
This article is protected by copyright. All rights reserved.
Accepted Article
This article is protected by copyright. All rights reserved.
