Journal of Environmental Management 166 (2016) 250e259

Contents lists available at ScienceDirect

Journal of Environmental Management

journal homepage: www.elsevier.com/locate/jenvman

Research article

Forest clearfelling effects on dissolved oxygen and metabolism in

peatland streams
Connie O'Driscoll a, b, *, Mark O'Connor a, b, Zaki-ul-Zaman Asam a, b, Elvira de Eyto c,
Lee E. Brown d, Liwen Xiao b
a
Department of Civil Engineering, National University of Ireland Galway, Galway, Ireland
b
Department of Civil, Structural and Environmental Engineering, Trinity College Dublin, Dublin 2, Ireland
c
Marine Institute, Newport, Co., Mayo, F28 PF65, Ireland
d
School of Geography/water@Leeds, University of Leeds, Leeds, LS2 9JT, United Kingdom

a r t i c l e i n f o a b s t r a c t

Article history: Peatlands cover ~3% of the world's landmass and large expanses have been altered significantly as a
Received 18 March 2015 consequence of land use change. Forestry activities are a key pressure on these catchments increasing
Received in revised form suspended sediment and nutrient export to receiving waters. The aim of this study was to investigate
12 October 2015
stream dissolved oxygen (DO) and metabolic activity response following clearfelling of a 39-year-old
Accepted 15 October 2015
Available online 26 October 2015
lodgepole pine and Sitka spruce forestry in an upland peat catchment. Significant effects of clearfelling
on water temperature, flows, DO and stream metabolic (photosynthesis, respiration) rates were revealed.
Stream temperature and discharge significantly increased in the study stream following clearfelling.
Keywords:
Stream metabolism
Instream ecosystem respiration increased significantly following clearfelling, indicating an increase in
Dissolved oxygen the net consumption of organic carbon.
Water quality © 2015 Elsevier Ltd. All rights reserved.
Temperature
Forest clearfelling
Blanket peat catchments

1. Introduction biodiversity and ecosystem functioning globally (Vo € ro

€ smarty et al.,
2010). Headwater streams constitute a large proportion of aquatic
Peatlands account for circa 3% of the Earth's total landmass (Bain systems and have the ability to affect a large percentage of fresh-
et al., 2011), yield 10% of the world's freshwater supply and water resources (e.g. Roberts et al., 2007). These streams are
comprise one-third of the global soil carbon (Joosten and Clarke, considered to have a far greater role in biogeochemical cycling than
2002). Peatland streams play a significant role in the global car- recognised previously (Benstead and Leigh, 2012) and can highlight
bon cycle and thus climate change, both by sequestering carbon land use impacts at the scale of first order streams which may be
and releasing it to the atmosphere (Billett et al., 2007). Peatland diluted at the catchment scale (O'Driscoll et al., 2013; Rodgers et al.,
area has significantly diminished since the 1800s due to climate 2010). Headwater peatland catchments contribute significantly to
change and land use management (Joosten and Clarke, 2002). the biological and genetic diversity of north-western European
Peatland conversion to forestry was commonly adopted in north- countries (Ramchunder et al., 2011; O'Driscoll et al., 2012; Drinan
western Europe, during the late 20th century (Paavilainen and et al., 2013). While there have been some detailed considerations
€iv€
Pa anen, 1995) with a view to improving an unexploited natural of how land use changes affect peatland stream ecosystems using
resource. These blanket peat forests have reached harvestable age indicators such as macro-invertebrates and diatoms (Brown et al.,
and concerns have been raised about the impacts of forestry 2013; O'Driscoll et al., 2012, 2014a; Ramchunder et al., 2009,
clearfelling on the receiving aquatic systems. 2012, 2013), few studies have examined the role of land use
Human land use has been identified as a major threat to aquatic change on instream metabolism, and the diurnal fluctuations of DO
(but see Young and Huryn, 1999).
In many northern European countries, coniferous trees are
* Corresponding author. Department of Civil Engineering, National University of currently harvested in sensitive peatland forest catchments, raising
Ireland Galway, Galway, Ireland. concerns about the potential impact on the receiving waters
E-mail address: connieodriscoll@gmail.com (C. O'Driscoll).

http://dx.doi.org/10.1016/j.jenvman.2015.10.031
0301-4797/© 2015 Elsevier Ltd. All rights reserved.
 C. O'Driscoll et al. / Journal of Environmental Management 166 (2016) 250e259 251

(Nugent et al., 2003). An estimated 500,000 ha of peatland was stream flow rates would increase; and; (H4) metabolic rates would
afforested between the 1950s and 1990s in the UK and 300,000 ha rise post-clearfelling due to increased nutrients, light and tem-
in Ireland (EEA, 2004; Hargreaves et al., 2003). Many studies have perature in the study stream. From a global perspective this study
reported catchment losses of suspended sediment and nutrients provides a unique opportunity to develop our understanding of the
following forest harvesting activities on peatland (e.g. Cummins impacts of forest clearfelling on stream functional ecology. This
and Farrell, 2003; Rodgers et al., 2010). More recently studies need for understanding is essential for enhancing the conservation
have begun to examine the pathways and mechanisms of nutrient of freshwater ecology in these habitats, informing management
and sediment release (Asam et al., 2012, 2014; O'Driscoll et al., practices and underpinning conservation schemes.
2014b). While some studies have reported a reduction in DO con-
centrations in aquatic ecosystems following forest clearfelling ac- 2. Methods
tivities on blanket peat (Drinan et al., 2013; Finnegan et al., 2014),
the potential causes and effects on ecological processes have not 2.1. Study sites
been investigated.
Dissolved oxygen (DO) is one of the most vital components of The study was carried out from March 2009 to January 2012 in
water quality in surface water bodies (Brooks et al., 1997). It is Glennamong, a 17.9 km2 sub-catchment of the Burrishoole catch-
essential for aerobic respiration at all trophic levels, particularly in ment, Co. Mayo (53
580 N, 9
370 E, 69 m a.s.l; Fig. 1). Catchment
headwater streams where some organisms (e.g. fish) may have topography is mountainous with a maximum elevation of 716 m.
high metabolic demands (Guignion et al., 2010). Diffusion from the The Burrishoole has a temperate oceanic climate due to its prox-
atmosphere at the stream surface exchange, mixing of the stream imity to the Atlantic coast with mean annual rainfall of
water at riffles, and photosynthesis from in-stream primary pro- 1560 mm year1 (McGinnity et al., 2009). The Glennamong catch-
duction provide the principal sources of in-stream DO. DO can ment was planted in 1972 with a combination of Lodgepole pine
become depleted when water bodies become stagnant leading to (Pinus contorta) (86%) and Sitka spruce (Picea sitchensis) (13%) using
increased consumption of oxygen by microbial organisms. spaced-furrow ploughing, creating furrows and ribbons (over-
Increased inputs of chemicals that react readily with oxygen in the turned turf ridges). The trees were planted on ribbons at 1.5 m
stream (reduction of nitrate (NO3) to ammonia (NH4)) can also intervals, giving an approximate soil area of 3 m2/tree. Ground
cause oxygen depletion. Temperature can affect DO concentrations mineral phosphate was applied at a rate of 28.3 g per tree and was
physically with higher solubility of DO observed for colder waters, spot-applied manually immediately after planting. Stand density
or indirectly via the significant role of temperature in ecosystem was reduced to ~2800 trees ha1 by natural die off before clearf-
metabolism (Yvon-Durocher et al., 2010). elling. The basal area for the stand was ~56 m2 ha1. The remainder
Alongside natural drivers of DO dynamics, stream DO concen- of the subcatchment is commonage (i.e. land that is owned by more
trations can also be affected by forest management activities. than one person) and is grazed extensively by sheep. Peat soil depth
Clearfelling may introduce brash material into receiving aquatic at the sites is > 1 m and overlies mainly quartzite and schist
systems, potentially increasing organic matter supply and thus bedrock.
biological oxygen demand (BOD) (Lockaby et al., 1997). Forest Small catchments drained by two first order streams that flow
clearfelling, site fertilisation and preparation can stimulate eutro- directly into the Glennamong stream were studied. Both catch-
phication via increased nutrient export to receiving waters, and ments are approximately 0.1 km2 in area and the study streams had
increased light availability and temperature following canopy a mean width of approximately 50 cm. Both streams largely flow
removal. Eutrophication generally promotes excessive plant over bedrock although some sections have a peat substratum. One
growth and decay, eventually causing a severe reduction in DO. catchment was clearfelled during the study period and is herein
Peatland soils are characterised by low density and can be easily referred to as Glennamong Study (GS). The second catchment
eroded in the absence of vegetation cover leading to increased received no management intervention during the study period and
suspended sediment export. Organic matter when present in sus- is herein referred to as Glennamong Control (GC; Fig. 1). Clearfelling
pended sediment is biologically active and as a consequence con- commenced in GS on February 8th 2011 and finished at the end of
tributes to the oxygen consumption in streams during March 2011. A harvester machine was used to clearfell the 9.4 ha GS
decomposition (Paavilainen and P€ €nen, 1995). Moreover, sus-
aiva catchment. Clearfelling at the GS was carried out in accordance
pended sediment might also reduce photosynthesis via reduced with best management practices (BMPs) (Forest Service, 2000a, b),
light penetration and bed smothering (Davies-Colley et al., 1992; as far as practicable (see Finnegan et al., 2014 for more detail). The
Van Nieuwenhuyse and LaPerriere, 1986). An additional strong crown of the tree and associated residues (i.e. needles, twigs and
driver of changes in DO concentrations is likely to be altered stream branches) were collected to form windrows and brash mats which
thermal regime, after canopy removal leads to significant increases were used for machine travel thus protecting the soil surface and
in net radiation (Hannah et al., 2008; Brown et al., 2010). reducing erosion. The windrows/brash mats (~4 m wide) were laid
The overarching aim of this study was to increase understanding parallel to the study stream and furrows on the harvested site,
of the effects of land use change due to forest clearfelling on which were at right angles to the contours. Surface water flowed
peatland stream ecosystems. DO concentrations were monitoring along the furrows and into collector drains that discharged into the
continuously over a two-year period in two first-order forested study stream.
headwater streams in Ireland, both with high gradient channels
and a bedrock/peat substrate. Commercial, non-native coniferous 2.2. Instrumentation and sampling
forestry was clearfelled from the catchment of one of these streams,
while the forestry was left intact in the catchment of the control GS and GC were instrumented at stable channel sections
stream. Based on the findings of earlier studies of forestry clearf- downstream of the forested areas in January 2010. H-flumes, a
elling effects on receiving streams (Drinan et al., 2013; Finnegan water level recorder (OTT SE200, Germany), a Datasonde (Hydro-
et al., 2014) we hypothesised that following clearfelling: (H1) lab, USA) measuring water temperature and DO and an OTT®
there would be significant decreases in DO concentrations; (H2) LogoSens 2 data logger (5 min resolution) were installed at both
periphytic biomass would increase in the stream due to increased stations. Data loggers were downloaded and Datasondes recali-
light, temperature and nutrients, (H3) both water temperatures and brated every four weeks. Periphyton samples were collected
252 C. O'Driscoll et al. / Journal of Environmental Management 166 (2016) 250e259

Fig. 1. Map of the Glennamong catchment showing locations of the control (GC) and study (GS) areas.

quarterly (March 2009eAugust 2011) from five cobble surfaces 2.3. Stream metabolism
using 100 ml of stream water (Biggs and Kilroy, 2000). Orthogonal
measurements of each stone were taken in the field, and converted Metabolism was calculated from measurements of diel tem-
to stone surface area (Dall, 1979). Samples were stored in the dark perature and DO concentration using the one-station open-channel
and analysed in the laboratory later the same day for chlorophyll a method (Odum, 1956). Metabolism was estimated by quantifying
(Chl a) and ash free dry mass (AFDM) (APHA, 2005). Simulta- changes in DO concentration at a single site over time and adjusting
neously, grab water samples were taken and analysed for total for DO exchange with the atmosphere (Bott, 2006). Flow duration
reactive phosphorus (TRP), ammonium-N (NH4eN) and nitrate-N curves were developed to distinguish between peak (Q5) and low
(NO3eN) colorimetrically using a nutrient analyser (Konelab 20, flow (Q95), and metabolism data collected during peak flows were
Thermo Clinical Labsystems, Finland). Suspended sediment (SS) omitted due to the potential for inaccuracies in the estimated
analysis was carried out by passing a known volume of water reaeration coefficient (DaSilva et al., 2013). Likewise for very low
through a pre-dried and weighed 1.2 mm GF/C filter disc (Whatman, flows the following rationale was adopted. Open channel flow
England) under suction. The filter and retained sediment were then nozzles (H Flumes) with shaft encoders were employed and were
dried at 105
C for 24 h and reweighed to give the total SS (APHA, rated under theoretical conditions to be operated at a minimum
2005). head of 0.005 m. However, to set the head for the device being used
Water samples were also taken as part of two complementary to record water level, a prerequisite is the need for water to be
studies on a monthly basis in the pre-felling study period and present in the flume, posing a potential source of error due to
weekly in the post-felling period using an automated ISCO sampler, turbulence. For this reason water level of 0.01 m was chosen as a cut
the findings of which are published in O'Driscoll et al. (2014b) and off point for limit of detection (negligible or no flows), to increase
Finnegan et al. (2014). Storm events were targeted as Rodgers et al. confidence in the flow data.
(2010) found that nutrient concentrations were low during base Ecosystem respiration (ER) and the reaeration coefficient (K)
flow conditions. The sampling frequency was dependent on the were estimated for each 24-h cycle using the night-time regression
duration of the storm event which was estimated using weather method (Hornberger and Kelly, 1975), for which the changes in DO
predictions (e.g. hourly samples in a 24 h event; every 2 h during a concentration per hour during the dark period of each 24-h cycle
48 h event and so on). O'Driscoll et al. (2014b) examined the effi- (0:00 h to sunrise and sunset to 0:00 h) were regressed against the
ciency of a buffer zone on ameliorating nutrient and suspended DO deficit so that ER (mg O2 m2 h1) is equal to the intercept and K
sediment export arising from clearfelling. Finnegan et al. (2014) (h1) is equal to the slope. Night-time periods that yielded re-
investigated the implications of applied best management prac- gressions with significant slopes (P < 0.05) were used for the
tice for peatland forest harvesting incorporating phosphorus, ni- calculation of ecosystem metabolism. Additionally, the diel pat-
trogen, suspended sediment, DO, electrical conductivity, pH and terns of DO were inspected visually for outliers and 24-h periods
stream water temperature. This study examines in more depth the were omitted where such data was present (Izagirre et al., 2007).
DO and temperature impacts arising from clearfelling by consid- Unlike empirical equations that are commonly used to estimate K,
ering a longer pre-felling period and investigating flow regimes and the night-time regression method allowed determination of K
metabolic processes. independently of discharge (Q), and thereby did not violate the
assumption of independence for the later examination of meta-
bolism variability as a function of Q using linear regression. ER20

and K20
were calculated from mean night-time temperature and
 C. O'Driscoll et al. / Journal of Environmental Management 166 (2016) 250e259 253

then adjusted for specific temperature at each time interval emphasised significant residual autocorrelation, thus generalised
following the methods of Erlandsen and Thyssen (1983) and least squares (GLS) regression was used (Pinheiro et al., 2006).
Thyssen et al. (1983) respectively. Daily rates were determined Models took the form
from the sum of hourly rates. Gross primary production (GPP) for
each time interval (dt) was calculated from the following equation: TGS ¼ a þ bTGC þ ß2 sin ð2pj=DÞ þ ß3 cos ð2pj=DÞ þ e (2)

GPP ðdtÞ ¼ dC=dt  K ðCs  CÞ þ ER þ A (1) where TGS ¼ water temperature at GS, a ¼ regression intercept,
b ¼ regression coefficient, TGC ¼ water temperature at GC,
where C is the concentration of DO at a given time, Cs is the satu- j ¼ calendar day of year, D ¼ number of days in year (i.e. 365), and
rating oxygen concentration, and ER is expressed as absolute e ¼ error term. Error terms were modelled as first order autore-
values. Accrual of ground water (A) was judged to be negligible gressive processes based on a priori examination of autocorrelation
(Izagirre et al., 2007). Daily GPP (mg m2 day1) was calculated and partial autocorrelation functions.
from the sum of GPP rates for each time interval during daylight. Regression models were used subsequently to predict water
The daylight period was calculated using times from a sunrise ⁄ temperature in the post-clearfelling phase. The approximate sta-
sunset calculator (http://www.timeanddate.com/) for Dublin, tistical significance of clearfelling impacts was assessed by calcu-
Ireland, and 20 min added for the west of Ireland. lating a measure of random disturbance (ût) (Gomi et al., 2006;
A caveat with all studies of whole stream metabolism is the Watson et al., 2001):
accuracy of the estimations of the metabolic parameters with po-
tential for error occurring throughout the stages of calculations b t ¼ ðyt  b
u y tÞ  r1 ðyt  1  b
y t  1Þ (3)
(Demars et al., 2011). Indeed, methods of assessment for ecosystem
functioning are still undergoing development (Aristegi et al., 2009; where y is the observed water temperature and y ^ is the predicted
Demars et al., 2011) and improvement. One of the main issues is the water temperature on day t, and r1 is the lag1 autocorrelation
quantification of the reaeration rate, but in a study comparing coefficient from the GLS regression. 95% confidence intervals of
several methods (Aristegi et al., 2009) the night-time method disturbance estimates were calculated as 1.96 (s ût). If there was no
proved the most robust and reliable among those tested. While the effect of forest clearfelling on water temperature, ût in the post-
reaeration rates estimated in our study were derived from this felling period would be similar to the pre-clearfelling period; this
method rather than being measured directly in situ (e.g. using gas hypothesis was tested with a two sample KolmogoroveSmirnov
additions), a consistent calculation and approach was applied to the test (Gomi et al., 2006). Linear mixed-effects (LME) models were
two sites pre- and post-felling data; thus, the relative magnitude of fitted using the library nlme (Pinheiro and Bates, 2000) to test the
effect between before-after and control-impact was considered significance of forest clearfelling on instream metabolism and pri-
reliable. mary productivity. This approach was adopted as there were
The volumetric rates of ER and GPP were converted to areal repeated measures taken at the same sites over time. The model
units (mg m2 day1) by multiplying by the mean reach depth (m). included Site (GS and GC) as a fixed effect and Treatment (before
Net ecosystem production (NEP) was the difference between GPP and after clearfelling) as a random effect. All statistical tests were
and ER, and P⁄R was GPP divided by ER. Negative ER is used to show performed with the R software package (R Development Core
oxygen consumption and a larger negative number indicates Team, 2008).
increased respiration. All metabolism calculations were performed
using the RIVERMET© spreadsheet package (Izagirre et al., 2007). 3. Results
This analysis resulted in 198 estimates of open-water metabolism
per year at GS and 224 at GC. There were a total of 83 concurrent 3.1. Meteorological observations
values for the two streams, 55 before clearfelling and 28 after.
The annual mean air temperature was 10.2
C during the study
2.4. Statistical analyses period, varying from a daily minimum of 7.5
C to a daily
maximum of 18.8
C. There was a significant difference between the
Comparisons of daily precipitation and air temperature in the pre- and post-clearfelled daily air temperature with higher tem-
years before and after clearfelling were made using repeated peratures observed in the pre-felling period (RM ANOVA,
measures (RM) analysis of variance (ANOVA) models. Flow and DO P < 0.0001) (Fig. 2a).
measurements were averaged by day to reduce the number of Total precipitation from a nearby Met Eireann weather station
observations and eliminate a falsely enhanced P value. Significance (circa 5.5 km south of the study site) was 1398 mm pre-felling
was determined at the 1% level unless otherwise noted. Repeated (February 2010eJanuary 2011) and 1981 mm post-felling
measures (RM) ANOVA models were used to test the significance of (February 2011eJanuary 2012) (Marine Institute, unpublished
site (GS and GC) and treatment (before and after clearfelling) in DO data) being significantly higher after clearfelling (RM ANOVA,
concentration, and flow. Datasets were ‘matched’ by retaining only P < 0.0001) (Fig. 2b). The number of wet days (>1 mm of precipi-
those days where data were available for both GC and GS. The data tation/day (Hundecha and Ba rdossy, 2005)) recorded was 174
were tested for normality and homogeneity of variance to ensure before clearfelling and 244 after clearfelling. The mean daily pre-
the assumptions of linear modelling were met. cipitation before clearfelling was 3.8 mm and 5.4 mm after
To assess whether clearfelling influenced stream thermal clearfelling.
regime, a generalised regression approach was used to predict daily
mean and maximum stream temperature at GS from measure- 3.2. Streamwater observations
ments at GC following the methods of Gomi et al. (2006) and
Dickson et al. (2012). Pre-clearfelling data was used to create Nutrient concentrations in GS were low pre-felling with mean
models, and subsequently to assess the impacts of stem abstraction concentrations of 12.6 ± 2.2 mg L1, 75.1 ± 45.4 mg L1, <50.0 mg L1
during clearfelling and in the immediate post-clearfelling phase. for TRP, NHþ4  N and NO3eN respectively (Table 1). Similarly, pre-
Primary exploratory ordinary least squares (OLS) regression, felling mean concentrations at GC were 14.6 ± 3.9 mg L1,
autocorrelation analyses and DurbineWatson statistics 71.4 ± 49.6 mg L1 and 67.2 ± 24.1 mg L1 for TRP, NHþ 4  N and
254 C. O'Driscoll et al. / Journal of Environmental Management 166 (2016) 250e259

Fig. 2. Time series of (a) daily air temperature, (b) total daily rainfall recorded at the Met Eireann weather station (circa 5.5 km south of the study site). Vertical dashed lines indicate
the timing of forest clearfelling.

Table 1
Nutrient and suspended sediment concentrations in the GC (control) and GS (clearfelled) reaches, in the pre- and post-felling periods. (The summary statistics as indicated
by ± represent the standard deviation. Nine storm events are represented in the pre-felling period and 18 in the post-felling period, with 24 water samples collected in each
event).

GC pre-felling GC post-felling GS pre-felling GS post-felling

TRP (mg L1) 14.6 ± 3.9 13.5 ± 5.30 12.6 ± 2.20 38.1 ± 23.6
NeNH4 (mg L1) 71.4 ± 49.6 47.5 ± 30.9 75.1 ± 45.4 69.3 ± 26.8
NeNO3 (mg L1) 67.2 ± 24.1 78.4 ± 63.7 <LOD 72.5 ± 51.4
SS (mg L1) 116 ± 258 9.00 ± 8.60 75.1 ± 52.0 34.3 ± 31.7
DO (mg L1) 11.3 ± 1.26 11.0 ± 0.94 10.5 ± 2.86 9.24 ± 2.84
N⁄ P 5.00 5.50 6.20 2.10

NO3eN respectively (Table 1). The GS TRP concentrations increased Mean stream discharges over the duration of the study period
two-fold from 12.6 ± 2.2 mg L1 to 38.1 ± 23.6 mg L1 and the GC TRP were 3.60 l s1 and 3.16 l s1at the GS and GC, respectively (Fig. 3b).
concentrations remained the same (Table 1). At GC the NHþ 4 N During the winter period, the mean low flow at GS increased 2-fold
concentrations almost halved from pre-to post-felling periods (from 2.29 to 4.68 l s1) whereas the GC observed a 1.5-fold in-
(47.5 ± 30.9 mg L1) whereas the GS pre- and post-felling periods crease (from 2.13 to 3.21 l s1). During the summer period, the
remained the same. The N: P ratio at GS shifted from 6.2 to 2.1after mean low flow at GS also increased 2-fold (from 2.26 to 4.71 l s1)
the onset of clearfelling (Table 1). whereas the GC observed an 1.5-fold increase (from 2.15 to
Daily mean DO concentrations ranged from 6.24 to 14.7 mg L1 3.12 l s1), a difference which was statistically significant (RM
at the GC and 0.05e14.8 mg L1 at the GS over the duration of the ANOVA, P < 0.0001).
study period (Fig. 3a; Table 1). ANOVA results indicated that Daily mean stream water temperatures in the pre-felling period
clearfelling significantly impacted DO (RM ANOVA, P < 0.0001) ranged from 0.14 to 15.9
C in the GC and 1.92e13.1
C in the GS
with lower values of DO concentrations observed in the GS post- (Fig. 3c). Post-felling stream temperatures ranged from 5.65 to
felling than the GC. The scatterplot of daily mean DO concentra- 14.9
C in the GC and 6.09e14.5
C in the GS. Mean (±St. dev) ût
tion from GC to GS for both pre- and post clearfelling showed that values for pre-felling were 0.007 (±0.32), but post-felling these
more days had lower average DO concentrations at both sites post- rose to 0.32 (±0.31), with the increase in ût post-felling being sta-
felling compared with pre-felling (Fig. 4). Of all the DO measure- tistically significant (KS: D ¼ 0.69; P < 0.00001) (Fig. 5).
ments pre-clearfelling, 1% recorded (5-min increments) at GC and
15% at GS were less than 80% saturation. Post-felling, 2% of recorded
3.3. Functional response to clearfelling
DO measurements at GC and 23% at GS were less than 80%
saturation.
Ecosystem respiration (ER) was greater than GPP for all readings
 C. O'Driscoll et al. / Journal of Environmental Management 166 (2016) 250e259 255

Fig. 3. Trend of daily averages of (a) DO concentration (mg L1), (b) stream temperature (
C) and (c) stream flow (l s1) at control (GC) and clearfelled (GS) sites. Vertical dashed
lines indicate the timing of forest clearfelling. Low DO occurred in the study catchment before clearfelling as well as afterwards. Gaps represent missing data.

recorded indicating both the GS and GC streams were heterotrophic

over the duration of the study period. Gross Primary Productivity
(GPP) rates were not significantly different at the GS site before and
after clearfelling (P ¼ 0.7169) (Table 2). Ecosystem respiration (ER)
was significantly different at the GS following clearfelling
(P < 0.001) (Table 2). Net ecosystem production (NEP) was signifi-
cantly different at the GS following clearfelling (P < 0.001) (Table 2).
Chlorophyll a (Chl a) concentrations ranged from 0.1 mg m2 to
42.9 mg m2 at the GS and 0.4 mg m2 to 11.7 mg m2 at the GC and
while the maximum Chl a values were observed in the post-felling
period at the GS site the interaction term site x treatment was not
significant (RM ANOVA, P > 0.05). Ash Free Dry Mass (AFDM)
ranged from 0.86 g m2 to 11.6 g m2 at the GS and 1.02 g m2 to
5.91 g m2 at the GC and the interaction term site x treatment was
significant (RM ANOVA, P ¼ 0.04).

4. Discussion
Fig. 4. Scatterplots of daily-averaged DO concentration (mg L1) at GS vs GC during
both pre- and post-felling periods. The results of this study advance previous findings from
256 C. O'Driscoll et al. / Journal of Environmental Management 166 (2016) 250e259

Fig. 5. Mean (±St. dev) random disturbance (ût values) pre-and post-felling over the duration of the study period. Vertical dash lines indicate the felling period. Random
disturbance measured the approximate statistical significance of clearfelling impacts determined using regression models. Dashed horizontal lines in plots of treatment effects
denote 95% confidence intervals of the pre-felling model.

Table 2
Gross primary production (GPP), ecosystem respiration (ER), and net ecosystem respiration (NEP) rates, and chlorophyll a (Chl a) and ash free dry mass (AFDM) concentrations
at GC and GS in the pre- and post-felling periods. (The summary statistics as indicated by ± represent the standard deviation. (There were a total of 83 concurrent values for the
two streams, 55 before clearfelling and 28 after).* indicates significant differences pre- and post-felling for the GS catchment.

GC pre-felling GC post-felling GS pre-felling GS post-felling

GPP (mg O2 m2 d1) 0.13 ± 0.17 0.23 ± 0.33 0.20 ± 0.25 0.21 ± 0.35
ER (mg O2 m2 d1)* 0.56 ± 0.97 0.84 ± 0.91 1.39 ± 1.26 3.46 ± 5.85
NEP (mg O2 m2 d1)* 0.44 ± 0.89 0.61 ± 0.66 1.19 ± 1.17 3.36 ± 5.63
Chl a (mg m2) 3.50 ± 6.78 6.02 ± 6.33 7.55 ± 10.6 17.2 ± 27.1
AFDM (mg m2)* 2.38 ± 3.14 2.61 ± 1.49 4.22 ± 3.95 9.45 ± 7.81

peatland forest clearfelling research (Drinan et al., 2013; Finnegan Drinan et al. (2013) where elevated readings of BOD and COD were
et al., 2014; O'Driscoll et al., 2013) by increasing our understand- observed in runoff water from the clearfelled GS catchment with
ing of the impacts on DO concentrations and associated drivers. the BOD exceeding guidelines for good status under the European
Water quality impacts arising from peatland management have Community Environmental Objectives (Surface Waters) Regula-
received increased attention over the past decade (Holden et al., tions 2009 (EC, 2009).
2007). Studies have focused on sediment, and nutrient export Removal of the canopy provides increased light to the stream
and bio-indicators such as macroinvertebrates (e.g. Brown et al., which could directly enhance algal periphyton biomass (Lam, 1981;
2013; Drinan et al., 2013; Ramchunder et al., 2011; Rodgers et al., Holopainen and Huttunen, 1998). Moreover, increases in phos-
2010, 2011). This study has provided the first detailed insight into phorus concentrations to greater than 30 mg l1 can trigger eutro-
how instream metabolic rates respond to forest clearfelling on phication in freshwaters (Boesch et al., 2001). However, while TRP
blanket peat. concentrations did increase significantly (>30 mg l1) in the impact
Periods of low DO occurred in both the GS and GC in the pre- stream after clearfelling, the increase was not reflected in Chl a
felling period. Pre-clearfelling, 1% (5-min increments) of the DO biomass therefore H2 was partly rejected. Due to the large export of
measurements recorded at GC and 15% at GS were less than 80% terrestrial organic C in peatland streams (Ryder et al., 2014), in
saturation. This is consistent with the findings of other studies (e.g. addition to the increased export following clearfelling (Drinan
Da Silva et al., 2013; Ice and Sugden, 2003). Headwater forested et al., 2013), aquatic primary production can be somewhat con-
peatland streams are further subjected to conditions that naturally strained by decreased light penetration (Arvola, 1984). However,
limit DO such as the peat substrata on the stream bed (Drinan et al., benthic AFDM biomass did significantly increase following clearf-
2013; Finnegan et al., 2014; O'Driscoll et al., 2013). Post-felling, 2% elling supporting H2. Increases in benthic AFDM were also noted by
of recorded DO measurements at GC and 23% at GS were less than Brown et al. (2013) in peatland catchments subjected to burning
80% saturation. DO saturation below 80% is considered to pose and this was attributed to the increased vulnerability of organic
environmental risk for aquatic life by the Irish EPA (Bowman, 2009). soils to physical erosion. The emerging consensus is the effect of
The before-after-control-impact (BACI) comparison of daily mean disturbance increases benthic organic matter in peatland stream
DO concentration over the entire study period shows that although systems (Brown et al., in press).
there were notable periods of low DO in the GS pre-felling, there Clearfelling was shown to influence the thermal regime of the
was a significant increase in the number of occurrences post-felling study stream with stream temperatures increasing significantly
thus supporting H1, that there would be significant decreases in DO post-felling and supporting H3. These findings align with previous
concentrations post-felling. The reductions in DO concentration studies (Gomi et al., 2006; Moore et al., 2005). Water temperatures
could be attributed to higher concentrations of organic suspended in the post-felling winter period were warmer than the pre-felling
sediment post-felling. Finnegan et al. (2014) reported higher winter period at both the GS and GC; however, temperatures at the
organic suspended sediment at the GS following clearfelling, and as GC exceeded the GS in the summer period pre-felling with the
the organic component is biologically active oxygen could be uti- reverse observed in the post-felling summer period. Canopy
lised during decomposition (Greig et al., 2007; Paavilainen and removal eliminated the shading effect of the trees naturally
Pa€iv€
anen, 1995; Rodgers et al., 2011). This theory is supported by implying a change in lighting conditions in the open stretches of
 C. O'Driscoll et al. / Journal of Environmental Management 166 (2016) 250e259 257

the impact stream following clearfelling. Solar radiation is the was laid over the study stream in dry conditions. It is likely that the
predominant contributor of energy for summer warming in oxidation of organic matter and nitrification of the decomposing
streams with no canopy (Bowler et al., 2012). Gomi et al. (2006) brash was the mostly likely cause of the observed DO minima post-
suggest riparian areas along streams protect the stream from felling. The immediate reduction in DO could have been caused by
increased thermal variability, with effects varying to some degree the increase in BOD and heterotrophic processes; however, the
with buffer width. However, many of the earlier afforested blanket open station method does not allow discrimination between BOD
peat catchments in Ireland and the UK were established without and COD therefore further work is necessary to unpick these drivers
any buffer areas, and trees were planted and harvested to the within peatland streams. Heterotrophic bacteria are the primary
stream edge (Broadmeadow and Nisbet, 2004). mediators in the process of terrestrially derived carbon degradation
In addition to canopy removal, alteration of stream discharge (Battin et al., 2003; Meyer, 1994) but limited work has been carried
could also impact on the stream thermal regime (Gomi et al., 2006). out elucidating these responses. Finnegan et al. (2014) recom-
Headwater streams can be shallow and experience low discharge mended that site inspections for harvesting plans are carried out
enhancing the opportunity for warming. The GS discharges during or immediately after a period of prolonged rainfall. Further
increased two-fold following clearfelling thus supporting H2. The research is warranted into the use of metabolism as an indicator of
increase was significant. This finding is similar to previous studies land management impacts in blanket peat catchments as it is one of
which attributed the increase to a reduction in evapotranspiration the most integrative ecosystem functions.
following tree removal (Robinson et al., 2003). Stream water tem-
perature is known to have a clear impact on the bio-physico- 5. Conclusions
chemical integrity of streams (Schlosser, 1995; Stott and Marks,
2000). This impact is expressed primarily through its regulation Peatland management operations in many northern European
of DO solubility in water (Horne and Goldman, 1994) and the locations are, on the whole, environmentally challenging due to
growth, metabolism, and respiration of aquatic organisms (Eckert, high annual precipitation, high soil water content, low ground-
1988). It has been previously reported that it takes several years bearing capacity and the typically ecologically sensitive nature of
for upland blanket peat sites to revegetate following clearfelling the receiving waters. This study has provided a detailed insight into
(O'Driscoll et al., 2011); however, it is not clear how the thermal the effects of blanket peat clearfelling on instream DO and associ-
regime recovers with recovering growth in vegetation in the ated drivers and processes. Clearfelling significantly reduced the
ensuing years. DO concentrations in the stream and this was most likely linked to
The stream metabolism rates observed for this study were changes in respiration owing to increases in stream temperature
within the range reported across numerous other streams (e.g. and nutrient export to the stream following clearfelling. The find-
Clapcott and Barmuta, 2010; Da Silva et al., 2013) with GPP ings of this study highlight the need to develop our understanding
consistently below ER (e.g. Marzolf et al., 1994; Young and Huryn, of whether natural revegetation establishment will enable recovery
1996; Demars et al., 2011). These findings provide some confi- of the stream thermal regime and associated biogeochemical pro-
dence in the use of the single-station open-channel method, and cesses to pre-clearfelling levels. Furthermore, quantifying riverine
complement earlier studies which report that headwater streams GPP and ER, and linking these to nutrient cycling under different
have higher rates of respiration than primary production and a flow, temperature and nutrient concentration conditions, is needed
larger export of terrestrial organic C can be expected from forested to provide baseline data to underpin assessment of management
peatland streams (Birkel et al., 2013). No significant impact of interventions and environmental change effects in peatland
clearfelling was observed on GPP, suggesting that heterotrophic systems.
processes dominated prior to felling and the increased light and
nutrients made available by the clearfelling did not alter this state.
Acknowledgements
ER, however, increased significantly following clearfelling. This
finding contrasts with Da Silva et al. (2013) who reported no impact
This study was funded by the SANIFAC (grant number
on stream metabolism following forest clearfelling. Clapcott and
RSF07 552) and Forsite (grant number 11/C/208) projects which
Barmuta (2010) found significant increases in the mean values of
were funded by the Department of Agriculture, Fisheries and the
all functional variables with clearfelled streams in comparison to
Marine under the STIMULUS Programme 2007e2013. The authors
the control catchments but reported that degree of response
wish to acknowledge the input of Coillte in allowing access to state
depended on the underlying geology. The decrease observed in NEP
forestry for this project, and for providing logistical support and
indicates an increase in the net rate of organic carbon consumed
advice. The authors gratefully acknowledge the assistance of the
(Young et al., 2008). Forest clearfelling can cause inputs of fresh
Marine Institute. Two anonymous reviewers provided insightful
brash into receiving waters (Campbell and Doeg, 1989; Lockaby
comments on the manuscript.
et al., 1997), stimulating heterotrophic processes (Clapcott and
Barmuta, 2010), and Drinan et al. (2013) reported elevated BOD in
References
the GS following clearfelling supporting this theory.
Ponce (1974) reported on the high demand for oxygen exerted APHA, 2005. Standard Methods for the Examination of Water and Wastewater.
by microbes associated with fresh brash and mentioned it can American Public Health Association, Washington.
rapidly deplete DO concentrations in receiving water. The rate of Aristegi, L., Izagirre, O., Elosegi, A., 2009. Comparison of several methods to calcu-
late reaeration in streams, and their effects on estimation of metabolism.
chemical release and oxygen demand is highest during the first two Hydrobiologia 635, 113e124.
weeks post-clearfelling, which coincides with the time in this study Arvola, L., 1984. Vertical distribution of primary production and phytoplankton in
when a sharp decline in DO occurred in the study stream. Drinan two small lakes with different humus concentration in southern Finland.
Ecography 7 (4), 390e398.
et al. (2013) also noted increased COD in the study stream imme-
Asam, Z., Kaila, A., Nieminen, M., Sarkkola, S., O'Driscoll, C., O'Connor, M., Sana, A.,
diately following clearfelling. Clearfelling was carried out in Rodgers, M., Xiao, L., 2012. Assessment of phosphorus retention efficiency of
accordance with BMPs with respect to harvesting plans, coupe size, blanket peat buffer areas using a laboratory flume approach. Ecol. Eng. 49,
timber landing areas, use of brash mats, as far as practicable 160e169.
Asam, Z., Nieminen, M., O0 Driscoll, C., O'Connor, M., Sarkkola, S., Kaila, A., Sana, A.,
(Finnegan et al., 2014). However, a site inspection highlighted that a Rodgers, M., Zhan, X., Xiao, L., 2014. Export of phosphorus and nitrogen from
brash mat used by the forwarder for carrying timber off the site, lodgepole pine (Pinus contorta) brash windrows on harvested blanket peat
258 C. O'Driscoll et al. / Journal of Environmental Management 166 (2016) 250e259

forests. Ecol. Eng. 64, 161e170. Department of the Marine and Natural Resources, Dublin. http://www.
Bain, C.G., Bonn, A., Stoneman, R., Chapman, S., Coupar, A., Evans, M., Gearey, B., agriculture.gov.ie/media/migration/forestry/publications/harvesting.pdf.
Howat, M., Joosten, H., Keenleyside, C., Labadz, J., Lindsay, R., Littlewood, N., Forest Service, 2000b. Forest and Water Quality Guidelines. Department of the
Lunt, P., Miller, C.J., Moxey, A., Orr, H., Reed, M., Smith, P., Swales, V., Marine and Natural Resources, Dublin. http://www.agriculture.gov.ie/media/
Thompson, D.B.A., Thompson, P.S., Van de Noort, R., Wilson, J.D., Worrall, F., migration/forestry/publications/waterquality.pdf.
2011. IUCN UK Commission of Inquiry on Peatlands. IUCN UK Peatland Pro- Gomi, T., Moore, R.D., Dhakal, A.S., 2006. Headwater stream temperature response
gramme, Edinburgh. to clear-cut harvesting with different riparian treatments, coastal British
Battin, T.J., Kaplan, L.A., Newbold, D., Hansen, C.M.E., 2003. Contributions of mi- Columbia, Canada. Water Resour. Res. 42, W08437.
crobial biofilms to ecosystem processes in stream mesocosms. Nature 426, Greig, S.M., Sear, D.A., Carling, P.A., 2007. A review of factors influencing the
439e442. availability of dissolved oxygen to incubating salmonid embryos. Hydrol. Pro-
Benstead, J.P., Leigh, D.S., 2012. An expanded role for river networks. Nat. Geosci. 5, cess. 21, 323e334.
678e679. Guignion, D., Dupuis, T., Teather, K., MacFarlane, R., 2010. Distribution and abun-
Biggs, B.J.F., Kilroy, C., 2000. Stream Periphyton Monitoring Manual. New Zealand dance of salmonids in Prince Edward Island streams. Northeast. Nat. 17,
Ministry for the Environment. National Institute of Water and Atmospheric 313e324.
Research. Hannah, D.M., Malcolm, I.A., Soulsby, C., Youngson, A.F., 2008. A comparison of
Billett, M.F., Garnett, M.H., Harvey, F., 2007. UK peatland streams release old carbon forest and moorland stream microclimate, heat exchanges and thermal dy-
dioxide to the atmosphere and young dissolved organic carbon to rivers. namics. Hydrol. Process. 22, 919e940.
Geophys. Res. Lett. 34, L23401. http://dx.doi.org/10.1029/2007GL031797. Hargreaves, K.J., Milne, R., Cannell, M.G.R., 2003. Carbon balance of afforested
Birkel, C., Soulsby, C., Malcolm, I., Tetzlaff, D., 2013. Modeling the dynamics of peatland in Scotland. Forestry 76 (3), 299e317.
metabolism in montane streams using continuous dissolved oxygen measure- Holden, J., Shotbolt, L., Burt, T.P., Chapman, P.J., Dougill, A.J., Fraser, E.D.G.,
ments. Water Resour. Res. 49, 1e16. Hubacek, K., Irvine, B., Kirkby, M.J., Reed, M.S., Prell, C., Stagl, S., Stringer, L.C.,
Boesch, D.F., Brinsfield, R.B., Magnien, R.E., 2001. Chesapeake Bay eutrophication: Turner, A., Worrall, F., 2007. Environmental change in moorland landscapes.
scientific understanding, ecosystem restoration, and challenges for agriculture. Earth Sci. Rev. 82, 75e100.
J. Environ. Qual. 30, 303e320. Holopainen, A.L., Huttunen, P., 1998. Impact of forestry practices on ecology of algal
Bott, T.L., 2006. Primary production and community respiration. In: Hauer, F.R., communities in small brooks in the Nurmes experimental forest area, Finland.
Lamberti, G.A. (Eds.), Methods in Stream Ecology, second ed. Academic, San Boreal Environ. Res. 3, 63e73.
Diego, CA, pp. 663e690. Hornberger, G.M., Kelly, M.G., 1975. Atmospheric reaeration in a river using pro-
Bowler, D.E., Mant, R., Orr, H., Hannah, D.M., Pullin, A.S., 2012. What are the effects ductivity analysis. J. Environ. Eng. Div. ASCE 101,729e101,739.
of wooded riparian zones on stream temperature? Environ. Evid. 1, 3. Horne, A.J., Goldman, C.R., 1994. Limnology, second ed. McGraw-Hill, Inc., p. 576
Bowman, J., 2009. New water framework directive environmental quality standards Hundecha, Y., B ardossy, A., 2005. Trends in daily precipitation and temperature
and biological and hydromorphological classification systems for surface waters extremes across western germany in the second half of the 20th century. Int. J.
in Ireland. Biol. Environ. 109B, 247e260. Climatol. 25, 1189e1202.
Broadmeadow, S., Nisbet, T.R., 2004. The effects of riparian management on the Ice, G., Sugden, B.D., 2003. Summer dissolved oxygen concentrations in forested
freshwater environment: a literature review of best management practices. streams of northern Louisiana. South. J. Appl. For. 27, 92e99.
Hydrol. Earth Syst. Sci. 8, 286e305. Izagirre, O., Bermejo, M., Pozo, J., Elosegi, A., 2007. RIVERMET: an excel-based tool to
Brooks, K.N., Ffolliott, P.F., Gregersen, H.M., Debano, L.F., 1997. Hydrology and the calculate river metabolism from diel oxygen-concentration curves. Environ.
Management of Watersheds. Iowa State University Press, Ames, IA. Model. Softw. 22, 24e32.
Brown, L.E., Johnson, K.L., Palmer, S., Holden, J., Ramchunder, S.J., Grayson, R., 2015. Joosten, H., Clarke, D., 2002. Wise use of Mires and Peatlands, a Framework for
Effects of fire on the hydrology, biogeochemistry and ecology of peatland river Decision Making. International Mire Conservation Group & International Peat
systems. Freshw. Sci. http://dx.doi.org/10.1086/683426. Society.
Brown, L.E., Johnston, K., Palmer, S.M., Aspray, K.L., Holden, J., 2013. River ecosystem Lam, W.Y.C., 1981. Ecological studies of phytoplankton in the Waikato River and its
response to prescribed vegetation burning on blanket peatland. PLoS One 8 catchment. N. Z. J. Mar. Freshw. Res. 15, 95e103.
(11), e81023. http://dx.doi.org/10.1371/journal.pone.0081023. Lockaby, B.G., Jones, R.H., Clawson, R.G., Meadows, J.S., Stanturf, J.A., Thornton, J.C.,
Brown, L.E., Cooper, L., Holden, J., Ramchunder, S.J., 2010. A comparison of stream 1997. Influences of harvesting on functions of floodplain forests associated with
water temperature regimes from open and afforested moorland, Yorkshire low-order, blackwater streams. For. Ecol. Manag. 90, 217e224.
Dales, northern England. Hydrol. Process. 24, 3206e3218. Marzolf, E.R., Molholland, P., Steinman, A.D., 1994. Improvements to the diurnal
Campbell, I.C., Doeg, T.J., 1989. Impact of timber harvesting and production on upstream-downstream dissolved oxygen change technique for determining
streamsdA review. Aust. J. Mar. Freshw. Res. 40, 519e539. whole stream metabolism in small streams. Can. J. Fish. Aquat. Sci. 51,
Clapcott, J.E., Barmuta, L.A., 2010. Forest clearance increases metabolism and 159e1599.
organic matter processes in small headwater streams. J. N. Am. Benthol. Soc. 29, McGinnity, P., Jennings, E., de Eyto, E., Allott, N., Samuelsson, P., Rogan, G.,
546e561. Whelan, K., Cross, T., 2009. Impact of naturally spawning captive-bred Atlantic
Cummins, T., Farrell, E.P., 2003. Biogeochemical impacts of clearfelling and refor- salmon on wild populations: depressed recruitment and increased risk of
estation on blanket peatland streams. I. Phosphorus. For. Ecol. Manag. 180 climate-mediated extinction. Proc. R. Soc. B Biol. Sci. 276, 3601e3610.
(1e3), 545e555. Meyer, J.L., 1994. The microbial loop in flowing waters. Microb. Ecol. 28, 195e199.
DaSilva, A., Jun Xu, Y., Ice, G., Beebe, J., Stich, R., 2013. Effects of timber harvesting Moore, R.D., Spittlehouse, D., Story, A., 2005. Riparian microclimate and stream
with best management practices on ecosystem metabolism of a low gradient temperature response to forest harvesting: a review. J. Am. Water Resour. Assoc.
stream on the United States Gulf coastal Plain. Water 5, 747e766. 41, 813e834.
Dall, P.C., 1979. A sampling technique for littoral stone dwelling organisms. Oikos Nugent, C., Kanali, C., Owende, P.M.O., Nieuwenhuis, M., Ward, S., 2003. Charac-
33, 106e112. teristic site disturbance due to harvesting and extraction machinery traffic on
Davies-Colley, R.J., Hickey, C.W., Quinn, J.M., Ryan, P.A., 1992. Effects of clay dis- sensitive forest sites with peat soils. For. Ecol. Manag. 180 (1e3), 85e98.
charges on streams: 1. Optical properties and epilithon. Hydrobiologia 248, O'Driscoll, C., de Eyto, E., Rodgers, M., O'Connor, M., Asam, Z.-Z., Xiao, L., 2013. Biotic
215e234. response to forest harvesting in acidic blanket peat fed streams: a case study

Demars, D.O.L., Manson, J.R., Olafsson, J.S., Gíslason, J.M., Gudmundsdo ttir, R., from Ireland. For. Ecol. Manag. 310, 729e739.
Woodward, G., Reiss, J., Pichler, D.E., Rasmussen, J.J., Friberg, N., 2011. Temper- O'Driscoll, C., Rodgers, M., O'Connor, M., Asam, Z.-Z., de Eyto, E., Poole, R., Xiao, L.,
ature and the metabolic balance of streams. Freshw. Biol. 56, 1106e1121. 2011. A potential solution to mitigate phosphorus release following clearfelling
Dickson, N.E., Carrivick, J.L., Brown, L.E., 2012. Flow regulation alters alpine river in peatland catchments. Water Air Soil Pollut. 221, 1e11.
thermal regimes. J. Hydrol. 464e465, 505e516. O'Driscoll, C., de Eyto, E., Rodgers, M., O'Connor, M., Asam, Z.-Z., Xiao, L., 2012.
Drinan, T.J., Graham, C.T., O'Halloran, J., Harrison, S.S.C., 2013. The impact of Diatom assemblages and their associated environmental factors in upland peat
catchment conifer plantation forestry on the hydrochemistry of peatland lakes. forest rivers. Ecol. Indic. 18, 443e451.
Sci. Total Environ. 443, 603e620. O'Driscoll, C., de Eyto, E., Rodgers, M., O'Connor, M., Asam, Z.Z., Kelly, M., Xiao, L.,
EC, 2009. Guidance Document No. 19 Guidance on Surface Water Chemical Moni- 2014a. Spatial and seasonal variation of peatland-fed riverine macro-
toring under the Water Framework Directive. European Union. invertebrate and benthic diatom assemblages and implications for assessment:
Eckert, R., 1988. Mechanisms and Adaptations. Animal Physiology. Freedman, New a case study from Ireland. Hydrobiologia 728, 67e87.
York, p. 683. O'Driscoll, C., O'Connor, M., Asam, Z.-Z., de Eyto, E., Rodgers, M., Xiao, L., 2014b.
EEA, 2004. Revision of the Assessment of Forest Creation and Afforestation in Creation and functioning of a buffer zone in an upland peat forest. Ecol. Eng. 62,
Ireland. Forest Network Newsletter Issue 150. European Environmental Agen- 83e92.
cy's Spatial Analysis Group. Odum, H.T., 1956. Primary production in flowing waters. Limnol. Oceanogr. 1,
Erlandsen, M., Thyssen, N., 1983. Modelling the community oxygen production in 102e117.
lowland streams dominated by submerged macrophytes. In: Lauenroth, W.K., Paavilainen, E., P€ aiv€
anen, J., 1995. Peatland Forestry: Ecology and Principles.
Skogerboe, G.V., Flug, M. (Eds.), Analysis of Ecological Systems: State of the Art Springer, Berlin.
in Ecological Modelling. Elsevier, pp. 855e860. Pinheiro, J.C., Bates, D.M., 2000. Mixed-effects Models in S and S-plus. Springer,
Finnegan, J., Regan, J.T., O'Connor, M., Wilson, P., Healy, M.G., 2014. Implications of New York, p. 528.
applied best management practice for peatland forest harvesting. Ecol. Eng. 63, Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D., 2006. Linear and Nonlinear Mixed Ef-
12e26. fects Models. Package nlme. Version 3.1e73. http://www.r-project.org/.
Forest Service, 2000a. Forest Harvesting and the Environment Guidelines. Ponce, S.L., 1974. The biochemical oxygen demand of finely divided logging debris
 C. O'Driscoll et al. / Journal of Environmental Management 166 (2016) 250e259 259

in stream water. Water Resour. Res. 10 (5), 983e988. Schlosser, I.J., 1995. Critical landscape attributes that influence fish population dy-
R Development Core Team, 2008. R: a Language and Environment for Statistical namics in headwater streams. Hydrobiologia 303, 71e81.
Computing. R Foundation for Statistical Computing, Vienna, Austria, 3-900051- Stott, T.A., Marks, S., 2000. Effects of plantation Forest clearfelling on stream tem-
07-0. http://www.Rproject.org. peratures in the Plynlimon experimental catchments, mid-Wales. Hydrol. Earth
Ramchunder, S.J., Brown, L.E., Holden, J., 2009. Environmental effects of drainage, Syst. Sci. 4 (1), 95e104.
drain-blocking and prescribed vegetation burning in UK upland peatlands. Prog. Thyssen, N., Erlandsen, M., Jeppesen, E., Holm, T.F., 1983. Modelling the reaeration
Phys. Geogr. 33, 49e79. capacity of low-land streams dominated by submerged macrophytes. In:
Ramchunder, S.J., Brown, L.E., Holden, J., Langton, R., 2011. Spatial and seasonal Lauenroth, W.K., Skogerboe, G.V., Flug, M. (Eds.), Analysis of Ecological Systems:
variability of peatland stream ecosystems. Ecohydrology 4, 577e578. State of the Art in Ecological Modelling. Elsevier, pp. 861e867.
Ramchunder, S.J., Brown, L.E., Holden, J., 2012. Catchment scale peatland restoration Van Nieuwenhuyse, E.E., LaPerriere, J.D., 1986. Effects of placer gold mining on
benefits stream macroinvertebrate biodiversity. J. Appl. Ecol. 49, 182e191. primary production in subarctic streams of Alaska. Water Resour. Bull. 22,
Ramchunder, S.J., Brown, L.E., Holden, J., 2013. Rotational vegetation burning effects 91e99.
on peatland stream ecosystems. J. Appl. Ecol. 50 (3), 636e648. Vo€ ro
€ smarty, C.J., McIntyre, P.B., Gessner, M.O., Dudgeon, D., Prusevich, A., Green, P.,
Roberts, B.J., Mulholland, P.J., Hill, W.R., 2007. Multiple scales of temporal variability Glidden, S., Bunn, S.E., Sullivan, C.A., Reidy Liermann, C., Davies, P.M., 2010.
in ecosystem metabolism rates: results from 2 years of continuous monitoring Global threats to human water security and river biodiversity. Nature 467,
in a forested headwater stream. Ecosystems 10, 588e606. 555e561.
Robinson, M., Cognard-Plancq, A.L., Cosandey, C., David, J., Durand, P., Führer, H.W., Watson, F., Vertessy, R., McMahon, T., Rhodes, B., Watson, I., 2001. Improved
Hall, R., Hedriques, M.O., Marc, V., McCarthy, R., McDonnell, M., Martin, C., methods to assess water yield changes from paired-catchment studies: appli-
Nisbet, T., O'Dea, P., Rodgers, M., Zollner, A., 2003. Studies of the impact of cations to the Maroondah catchments. For. Ecol. Manag. 143, 189e204.
forests on peak flows and baseflows: a European perspective. For. Ecol. Manag. Young, R.G., Huryn, A.D., 1996. Interannual variation in discharge controls
186, 85e97. ecosystem metabolism along a grassland river continuum. Can. J. Fish. Aquat.
Rodgers, M., O'Connor, M., Healy, M.G., O'Driscoll, C., Asam, Z., Nieminen, M., Sci. 53, 2199e2211.
Poole, R., Müller, M., Xiao, L., 2010. Phosphorus release from forest harvesting Young, R.G., Huryn, A.D., 1999. Effects of land use on stream metabolism and
on an upland blanket peat catchment. For. Ecol. Manag. 260 (12), 2241e2248. organic matter turnover. Ecol. Appl. 9 (4), 1359e1376.
Rodgers, M., O'Connor, M., Robinson, M., Müller, M., Poole, R., Xiao, L., 2011. Sus- Young, R.G., Matthaei, C.D., Townsend, C.R., 2008. Organic matter breakdown and
pended solid yield from forest harvesting on upland blanket peat. Hydrol. ecosystem metabolism: functional indicators for assessing river ecosystem
Process. 25, 207e216. health. J. N. Am. Benthic. Soc. 27, 605e625.
Ryder, L., de Eyto, E., Dillane, M., Poole, R., Jennings, E., 2014. Identifying the role of Yvon-Durocher, G., Jones, I., Trimmer, M., Woodward, G., Montoya, J.M., 2010.
environmental drivers in organic carbon export from a forested peat catchment. Warming alters the metabolic balance of ecosystems. Philos. Trans. R. Soc. B
Sci. Total Environ. 490, 28e36. Biol. Sci. 365 (1549), 2117e2126.