FORESTRY & SURFACE WATER ACIDIFICATION
(FORWATER)
2008
MARY KELLY-QUINNI, R OBERT C RUIKSHANKSI, JAMES JOHNSONI ,
RONAN MATSONI, JAN-R OBERT BAARSI & MICHAEL BRUEN2,
1
–School of Biology and Environmental Science
2 - School Of Architecture, Landscape & Civil Engineering
TABLE OF CONTENTS
1.
2
2.1
2.2
2.3
2.4
2.5
2.6
2.7
3
3.1
3.2
3.3
4
5
6
2
SUMMARY
INTRODUCTION
4
The source of the Pressure
Pathway Susceptibility
Receptor Sensitivity
5
6
6
SITE SELECTION & METHODOLOGY
8
SITE SELECTION
HYDROCHEMISTRY
MACROINVERTEBRATES
FISH
GEOGRAPHICAL IMAGING & SITE CHARACTERISATION
STATISTICAL ANALYSES
QUALITY CONTROL
8
11
13
14
17
18
20
RESULTS
21
HYDROCHEMISTRY
Conductivity
PH
Alkalinity, Sodium Dominance Index and Aluminium
Sources of Acidity
Comparison of Source Streams – Upstream and Downstream of
Forestry
MACROINVERTEBRATES
Community Composition
Selection of Macroinvertebrates Metrics
Taxon Richness
Ephemeropteran Richness
Ephemeropteran Abundance
EPT Metric
BMWP & ASTP Metrics
Community Diversity and Species Evenness
Clustering of Biological, Chemical and Physical Metrics
Multivariate Analyses
Evaluation of the Degree of Biological Impairment
Comparison of Source and Downstream Communities
Longitudinal Variation in Macroinvetebrate Metrics
Further Evaluation of Potential Longitudinal Patterns in
Macroinvertebrate Recovery from Forest Effects
FISH
21
21
22
28
32
34
DISCUSSION
REFERENCES
APPENDIX A
37
37
38
40
41
44
46
48
50
50
53
55
56
57
58
62
67
71
76
1
SUMMARY
This project set-out to investigate the presence and extent of acidification associated with
coniferous forestry in Ireland and to assess the risk of impact with respect to different
geological settings. In the selection of forested sites it was aimed to represent a combination
of the risk factors in terms of catchment forest cover and acid-sensitive geology that were
perceived to have the greatest potential for acidification. This study was designed to allow
comparisons of the hydrochemical and ecological quality of two groups of sites, forested and
non-forested, control or reference sites in two geological settings (igneous/metamorphic and
sedimentary) with four dominant soil types (peat, podzolic/lithosolic, poorly drained gleys
and well drained mineral soil). The 239 control and forested sites were categorised to reflect
a gradient in catchment forest cover (control (<5% forest cover) – 73 sites, and three
coniferous forestry bands (5-25, 25-50 and >50% coniferous cover with 27, 41 and 98 sites,
respectively).
Water samples were collected from all sites on three separate dates and
covered a range of flow conditions. Macroinvertebrate samples were collected in spring 2007
and electrofishing was carried out at 19 paired sites in summer 2007.
The pH results analyses suggested that most of the streams were episodically acidic with a
small group more likely to be circum-neutral. Overall, the pH results indicated increased
acidity at some sites associated with forestry on peat and podzolic/lithosoilic soils on both
igneous/metamorphic and sedimentary geology and to some extent on poorly drained gleys.
Furthermore, the frequency of low pH readings was substantially higher among some groups
of forested sites than the control sites.
Certainly the minimum pH for both peat and
podzolic/lithosolic sites on igneous/metamorphic geology began to fall below the lower limit
of the control sites when forest cover exceeded values in the region of 25-30%. The same
applied to peat sites draining sedimentary geology. Sites on podzolic/lithosolic soils on
sedimentary geology did not have minimum pH values below the lower limit of the control
sites until forest cover exceeded 60%. A similar threshold might be applied to sites on poorly
drained gleys but the level of replication was too low for this decision.
The presence of
forestry tended to depress site pH and alkalinity. Calculations suggested that dilution makes
a variable contribution to loss of alkalinity and in many cases the forested sites showed a
slightly higher % value. Anion titration was detected in all events examined. The principal
contributors were organic acids and excess sulphate, particularly in the east.
2
Overall, the biological data largely mirrored the trends for the acidity variables. Several
macroinvertebrate metrics (taxon richness, ephemeropteran richness, abundance of baetids,
EPT richness, diversity indices), which showed a strong relationship with pH, were also
shown to vary significantly across the forest cover bands or to correlate with % forest cover.
The analyses on the individual metrics highlighted similar % forest thresholds for risk of
impact as described for the hydrochemistry.
When a selection of non-correlated metrics
were combined it was clear that a large proportion of sites in the >50% cover band, and a
smaller number of the 25-50% band, had some degree of impairment.
However, not all
forested sites were impaired and further research must target these sites to better understand
the mechanisms governing responses to acid impact. Finally, the length of stream impacted
by forest-mediated acidification is likely to vary depending on geological and other
catchment characteristics.
The fish analyses was limited to 19 paired sites with similar habitat but did highlight
significant differences in fish catch and density between the control and forested groups. This
difference was mainly attributed to low numbers of fry (salmon and trout) in the forested
streams.
In terms of identified knowledge gaps we need to determine the acidification risk associated
with each of the key forestry practices from site preparation to felling. More detailed spatial
and temporal analyses of the chemical characteristics of waters draining sedimentary geology
is required for more precise mapping of acid sensitivity. The contribution of organic acids to
the acid pulses in control and forested sites on both igneous and sedimentary geology and the
process contributing to their release needs to be further explored. The influence of flowpath
and drainage on buffering potential and its interaction with acid inputs also requires further
research.
Finally, with respect to the biota it is critical to understand seasonal and
longitudinal changes in the community in response to acid inputs together with the functional
significance of any impairment.
3
1 INTRODUCTION
The Water Framework Directive (WFD) (2000/60/EC), which came into force in December
2000, requires EU member states to implement the necessary measures to prevent
deterioration in the status of all bodies of surface and groundwater and where necessary,
restore all waters to good ecological and chemical status by 2015.
As part of the
characterization process, the first risk assessment of the anthropogenic pressures on water
resources was undertaken to identify the pressures present in each river basin and the threat
they pose to the chemical and ecological status of waterbodies. The resulting “National
Characterisation Report for Ireland” (Anon, 2005) identified forestry (SD4) as one of the
land-use activities posing a potential risk in terms of diffuse pollution. Among the pressures
highlighted as arising from forestry were increased acidification from plantations in acidsensitive catchments (SD4a), sedimentation from clear fell, harvesting, new plantations, road
construction and erosion on steep catchments (SD4b) and eutrophication from fertilisation on
steep catchments and forest harvesting on peat soils (SD4c).
The Western River Basin District was given, inter alia, the task of further characterisation of
the risks from plantation forests and forest related activities on surface waters and to
subsequently developing a programme of measures to address any significant risks. The
present project was commissioned to address some of the knowledge gaps pertaining to the
acidification risk. The keys questions addressed by this research were:
1. Is there any evidence for a forest effect on hydrochemistry and biology?
2. What is the impact on the aquatic biota?
3. What conditions pose the greatest risks
The research sought to identify patterns that indicate risk and inform a ‘precautionary
approach’ through the first programme of measures.
The work was prefaced with a literature review which clarified the state of knowledge on
forest-mediated acidification. The literature review (Johnson et al. unpublished) followed the
Source–Pathway–Receptor model of the risk assessment and the key findings are summarised
here.
4
The Source of the Pressure
The role of forestry in the acidification of surface waters is primarily attributed to the
interception of atmospheric pollutants coupled with the inability, in sensitive areas, of the
substrate soils and geology to buffer the acidity (Jenkins et al., 1990; Ormerod et al., 1991).
The magnitude of the pressure exerted by the scavenging effect of forests depends primarily
on (a) the pollutant load and (b) the percentage of catchment forest cover. The extent of the
pressure is likely to vary with tree species with some species, such as Sitka spruce (P.
sitchensis), being more effective scavengers of pollutants than others. The pollutant load at a
site is further influenced by emission levels, climatic conditions such as the frequency and
magnitude of rainfall events, the amount of annual rainfall, prevailing wind direction and air
mass circulation patterns as well as site characteristics such as elevation and aspect, tree
species, stand age and structure. Few studies have attempted to clarify the relationship
between the extent of catchment afforestation and surface water chemistry across a range of
catchment characteristics but there is general agreement in the literature that an increase in
forest cover has the potential to increase the acidification pressure. However, as yet there is
no guidance on the threshold above which adverse impacts are apparent on stream chemistry
or biology in acid-sensitive areas.
Sea salt driven acid pulses can occur in coastal catchments. Forests capture marine ions as
wet or dry deposition (Farrell, 1995; Harriman, Anderson and Miller, 1995). During storm
events, high inputs of Na+ can displace other cations. The associated chloride ion is largely
conservative and most of it is quickly leached. As it passes through acidic soil to associated
drainage water it can be accompanied by H+ and Al3+. Other processes with potential to
contribute to acidification include (a) uptake of base cations by trees and subsequent removal
by harvesting, (b) oxidation and mineralization of organic matter producing organic acids and
(c) alterations to site hydrology resulting in the reduced residence time of water and (d) the
short-term release of nitrate following the large-scale felling of forest sites in acid-sensitive
catchments. Certainly increased drying of soil and altered drainage increases the oxidation of
organic matter and generates carboxylate anions, increases mineralization of organic matter
and consequently potentially increases in losses of ammonium and/or nitrate as well as
sulphate to drainage waters (Hornung et al., 1995). However, the contribution of these to
acidification processes in acid-sensitive areas has not been adequately assessed.
5
Pathway Susceptibility
The pathway susceptibility is primarily controled by catchment geology.
The available
literature indicates that water bodies susceptible to acidification are located in catchments
dominated by slow weathering bedrock such as granite and quartzite with shallow carbonate
free soils as well as areas of sandy, siliceous soils and highly weathered old leached soils
(Hornung et al., 1990). In Ireland, granitic areas in the west and northwest (Allott et al.,
1990; Bowman, 1991; Allott et al., 1997) as well as the east have been shown to be acidsensitive (Kelly-Quinn et al., 1996a; Kelly-Quinn, Tierney & Bracken, 1997). The potential
for acidification on Old Red Sandstone is likely to be less but this is not fully established and
is addressed in the current project.
Catchment size and hydrology/drainage also have a bearing on the susceptibility of running
waters to acidification. High drainage rates and steep topography of small upland catchments
reduces the contact time for runoff with bedrock and soil and consequently the time for soils
to impart buffering capacity to the runoff water (Waters and Jenkins, 1992). As a result,
waters draining smaller catchments may be more acidic and have higher concentrations of
inorganic monomeric aluminium reflecting a higher proportion of runoff from the acidic
mineral soils in the catchment. In larger catchments, the overall residence time of water in
soil system is longer and it is therefore likely to be more effectively neutralized.
The Sensitivity of the Receptor
The sensitivity of the receptor shows as an increases in acidity and in many cases elevated
aluminium concentrations (Ormerod et al., 1991).
The response of the biota to forest-
mediated acidification has been well documented (examples Clenaghan et al., 1998;
Harriman and Morrison, 1982; Stoner, Gee & Wade, 1984; Ormerod et al., 1991, Ormerod &
Wade, 1990; Allott et al., 1997; Tierney, Kelly-Quinn & Bracken, 1998). Biological impacts
associated with acidification in streams include 1) reductions in or total elimination of fish
populations, 2) reductions in taxon richness and elimination of some acid-sensitive
macroinvertebrate groups (particularly the Ephemeroptera) and 3) changes in the quality of
primary producers (Stoner, Gee & Wade, 1984; Ormerod, Wade and Gee, 1987; Rees and
Ribbens, 1995; Tierney, Kelly-Quinn & Bracken, 1998). The studies carried out in Ireland to
date have highlighted some impact in areas of Wicklow (Tierney, Kelly-Quinn and Bracken,
1998) and Galway and south Mayo (Allott et al., 1997) but no acidification-related impacts of
aquatic fauna were detected for sites on Old Red Sandstone in Cork (Giller et al., 1997).
6
However, the latter authors noted that some macroinvertebrate communities at medium
altitude (200-300m) with medium to high levels of forestry (25 to> 50%) seemed to resemble
communities at higher altitudes (>300m) than sites with low levels of forest cover at a
medium altitude. The present study set out to undertake more extensive sampling of Old Red
Sandstone to further assess the potential for acidification impacts. A further issue arising
from the AQUAFOR and indeed more recent WATERAC projects was that the occurrence of
detectable impact in rivers (as evidenced by loss of macroinvertebrate taxa and salmonids)
appeared to be rather patchy. Therefore, the current study set out to target good numbers of
sites with a combination of perceived high risk factors, so that a better measure of the extent
of impact could be achieved.
7
2. SITE SELECTION AND METHODOLOGY
2.1 SITE SELECTION
This study was initially designed to allow comparisons of the hydrochemical and ecological
quality of two groups of sites, forested and non-forested, control or reference sites. This
approach was adopted following consultation with the Forest and Water National Steering
Committee members (EPA, Cóillte, Irish Forest Service, Marine Institute, Central Fisheries
Board and National Parks and Wildlife) and is the current approach required by the WFD and
widely applied in freshwater studies. The forested sites were to represent a combination of
the risk factors perceived to have the greatest potential to facilitate acidification impacts.
Percentage forest cover in the catchment and geographical location were considered to
influence the magnitude of the acidification pressure. Factors influencing pathway
susceptibility were geology and soil type. A total of 239 sites (Appendix A), both control and
forested, were therefore selected to provide wide geographical coverage within acid-sensitive
geologies (igneous/metamorphic geology and Old Red Sandstone) and to encompass
combinations of geology and soil type (peaty and mineral). Four categories of soil were
targeted, well drained mineral, poorly drained gleys, podzolic-lithosols and peats [categories
followed consultation with Cóillte, Irish Forest Service and WRBD]. The forested sites were
selected to have >25% catchment forest cover, the majority of which was closed canopy
representing mature forests in the forest cycle. Catchment is here defined and applied
throughout this study as the drainage basin to the study site, alternatively referred to as site
watershed. Control sites were chosen within each of the regions where the forested sites were
located. To ensure comparability, where possible control and forested sites were within the
same larger river catchment, although adjacent catchments were selected in some instances.
However, the geographic spread of the control and forested catchments were similar.
Control sites were initially chosen to represent catchments with no forestry. However, during
the course of the study access to updated forest inventory information revealed variable
amounts of forest in some of the control catchments. Consequently, all 239 control and
forested sites were re-categorised to reflect the gradient in forest cover (control (<5% forest
cover) – 73 sites, and three coniferous forestry bands (5-25, 25-50 and >50% coniferous
cover) with 27, 41 and 98 sites, respectively). The numbers of sites in each forestry/soil/forest
cover category are given in Table 1 and their location is indicated in Fig. 1. Broadleaf
forestry represented a small percentage of the total forest cover and was not considered in the
8
analyses. Therefore, the forest cover values referred to throughout this report represent
coniferous plantation. There was no significant land-use pressure in the control catchments
apart from some rough grazing, although historical influences could not be ruled out entirely.
Table 1: Numbers of site sampled in each geology/soil category/forest cover category
Geology/Soils
Igneous/Metamorphic
Forest Cover
<5%
5-25%
25-50%
>50%
Totals
Peat
22
9
6
17
54
Podzolic/Lithosolic
11
4
9
8
32
Peat
20
4
11
30
65
Podzolic/Lithosolic
13
2
2
22
39
Gleys
6
3
7
11
27
Well Drained Mineral
2
5
6
9
22
Totals
74
27
42
97
239
Sedimentary
In each region the sites were selected on first to third order streams. A number of additional
sites were located further downstream to examine longitudinal changes in hydrochemistry
and aquatic biota. Every effort was made to control for slope, elevation and catchment size
assessed using channel length and stream complexity as indicated on OSI maps (1:50000)
and by restricting the majority of the sites to sub-catchments. Access to the sites was
facilitated by Coillte/Irish Forest Service, many of which were remote with the only access
by foot.
Each sites was represented by a 50 (macroinvertebrate sampling) to100 (fish sampling) metre
stretch. Physical descriptions for each of the sites were derived from GIS and from onsite
measurements of stream width (four measurements), depth (four measurements taken in
randomly selected pools in the stretch), substrate composition, mesohabitat cover and flow
condition (low, elevated and flood). Substrate was estimated as the percentage cover of
bedrock, boulder (26-200cm), cobble (6-26cm), gravel (0.4-6cm), sand (0.06-0.2cm) and silt
9
(<0.06cm). Mesohabitats were assigned to three categories, including riffles, glides and
pools.
• Control Sites
• 5-25% Forested
• 25-50% Forested
• >50% Forested
Fig. 1: Distribution of hydrochemical and macroinvertebrate sampling sites (colour coded for
forest-cover categories).
10
2.2 HYDROCHEMISTRY
Water samples were collected from all sites in clean one-litre and 250ml polypropylene
bottles (pH). Readings of conductivity (µS/cm), and oxygen (% saturation and mg/l O2) were
taken on site using automatic field meters and probes. All samples were sent to the Aquatic
Services Unit at the Environmental Research Institute (ERI, UCC) for analysis within 24hours of collection. A full suite of hydrochemical analyses were carried out using the
methodologies outlined in Table 2. Three rounds of water sampling were undertaken, 2/5/076/6/07, 13/11/07-21/12/07 and 18/3/08-16/4/08. The aim was to sample each site at variable
flow conditions, from low flow to flood.
samples for all sites.
It was however not possible to obtain flood
Furthermore, it was often difficult to ascertain the stage in the
hydrograph represented on any one date and therefore flow condition was simply recorded as
low, elevated or flood. Additional samples were taken upstream and downstream of the
forestry block on selected source streams. Samples were collected at comparable locations
on control streams at similar distances from the source (as indicated on OSI maps).
Sources of acidity and those potentially responsible for any observed differences between
forested and control sites were evaluated by examination of the results for sulphate, nitrate,
chloride and organic carbon. The potential loss of alkalinity during elevated flow due to
dilution by precipitation was assessed using the following formula applied by Kowalik et al.
(2007):
Alkalinity Dilution = ((( BC low - BC high)/ BC low) Alk. low)
____________________
______________________________________________________
(Alk. low - Alk. high)
BC=Base Cations, Alk.=Alkalinity
All concentrations are entered in units of µeq/l. High percentage values close to 100%
indicate that dilution is strongly affecting buffering.
Lower values indicate reduced
likelihood of dilution and possible titration by an acid anion.
11
Titration Ratio
Loss of alkalinity due to titration by an acidic anion is evidenced by changes in the following
ratio.
Titration Ratio = Alkalinity / BC
This can be confirmed with the results from the titration ratio equation. The proportional
contribution of acid anions to any titration processes was calculated as Anion/ Acid Anions
(Kahl et al., 1992). Losses in ANC due to sea salt effects were evaluated from changes in the
concentrations of Cl- and Na+ between low and high flow as applied by Evans et al (2008).
Table 2: Methods applied in the chemical analyses
Parameter
Method
Unit
pH
Conductivity
Alkalinity
Total Hardness
Colour
WTW pH330i pH meter
WTW LF330 Conductivity meter @ 25°C
Gran Titration
ETDA Titration
Colorimetric method using platinum/cobalt solution
as colour standard
SHIMADZU TOC-VCPH TOC analyzer
Automated Molybdate method using Lachat™
Quikchem FIA
Manual molybdate method after sample digestion
Automated salicylate method using Lachat™
Quikchem FIA
Automated colourimetric method using Lachat™
Quikchem FIA after cadmium reduction
Subtraction nitrite from TON
Manual colourimetric method
Graphite furnace AAS
Graphite furnace AAS after Amberlite™
Resin fractionation
Automated IC method using Lachat™ Quikchem IC
Automated IC method using Lachat™ Quikchem IC
Automated IC method using Lachat™ Quikchem IC
Automated IC method using Lachat™ Quikchem IC
Automated IC method using Lachat™ Quikchem IC
Automated IC method using Lachat™ Quikchem IC
Gravimetric method after filtering through
GF/C filter paper and dried at 104°C
Manual colourimetric method
Calculated
µS/cm @ 25ºC
mg/l CaCO3
mg/l CaCO3
Hazen Units
Dissolved Total Organic Carbon
Soluble Reactive Phosphorus
Total Phosphorus
Ammonia
Total Organic Nitrogen
Nitrate
Nitrite
Total Monomeric Aluminium
Inorganic Aluminium
Calcium
Magnesium
Potassium
Sodium
Chloride
Sulphate
Suspended Solids
Silicate
Sodium Dominance Index (SDI)
mg/ l DTOC
mg/l SRP
mg/l TP
mg/l Ammonia
mg/l TON
mg/l Nitrate
mg/l Nitrite
µg/l Aluminium
µg/l Aluminium
mg/l Ca2+
mg/l Mg2+
mg/l K+
mg/l Na+
mg/l Clmg/l SO42+
mg/l SS
mg/l Si
%
12
2.3 MACROINVERTEBRATES
Benthic macroinvertebrate samples were collected over a six week period in 2007 from the
beginning of April until the second week in May at the 239 sites. Additional samples were
collected on source streams above and below forest blocks and at similar points on
comparable streams.
A multi-habitat sampling approach was employed involving kick
samples of 1-minute duration taken using a standard pond net (mesh – 1mm). The time spent
sampling each mesohabitat (riffle, pool glide) was proportional to its percentage
representation in the study site ( Wright, 1995). Habitats contributing less than 5% of the
stable habitat in the reach were not sampled (Barbour et al., 1997). An additional one minute
was spent carrying out hand searches for attached invertebrates. Sampling was initiated
downstream of the reach and proceeded upstream. To avoid the confounding effects of
shading the forested sites were downstream of the forest within open, un-shaded reaches at
least 20m downstream of the forest block. Six samples were collected at each site and
preserved using 70% alcohol (IMS). These were sorted in the laboratory and the
macroinvertebrates were removed and identified to the lowest taxonomic unit possible using
FBA keys (Table 3). Identified samples were stored in 70% alcohol (IMS).
Table 3: Level of identification for macroinvertebrate groups.
Taxon
Plecoptera
Ephemeroptera
Trichoptera
Coleoptera
Chironomidae
Simuliidae
Other Diptera
Odonata
Hemiptera
Mollusca
Hirudinea
Oligochaeta
Level of Taxonomic Identification
Species
Species
Genus/species
Genus/species
Subfamily
Genus/
Family/genus/species
Genus/species
Genus/species
Species
Species
Order
13
2.4 FISH
Streams were selected on a paired catchment basis (one non-forested catchment, one forested
catchment) to represent similar physical characteristics including catchment area, elevation
and slope. With few exceptions paired streams were selected on the same main channel, in
close proximity so that both streams had the same source fish population within the main
channel. Fig. 2 illustrates this approach. In total, 38 sites were fished (19 non-forested and
19 forested sites) (Table 4, Fig. 3). Each site was fished using backpack electrofishing
equipment (Safari Research 550D backpack model). Site habitat characteristics including
numbers of riffles, glides and pools, stream width (four measurements) and pool depth (four
measurements) were recorded on site, along with on-site measurements of oxygen and
conductivity. The time taken to fish each site was recorded to compute fishing effort so that
comparisons could be made between forested and non-forested sites and corrected if
necessary for differences in effort.
A single-pass approach was adopted for the electrofishing sampling. Several investigations
have evaluated the accuracy and usefulness of single-pass electrofishing to estimate
abundance or relative abundance of salmonids in streams. These studies have indicated that
there is a significant relationship between number of fish caught in the first pass and the total
population size estimated from three or more passes (Hayes & Baird, 1994; Jones and
Stockwell, 1995; Kruse, Hubert & Rahel, 1998; Mitro and Zale, 2000; Arnason, Antonsson &
Einarsson, 2005; Bertrand, Gido & Guy, 2006) and it is therefore a sensitive method for
detecting differences in relative abundance. The FAME protocol recommends at least 10-20
times the wetted width be fished (Economou et al., 2002). As the majority of the selected
sites were approximately 2m wide, the 100m stretch fished in the present study was more
than adequate to satisfied this condition. To avoid the problem of shading sampling in the
forested catchments took place outside of the forest, usually immediately downstream (circa
20m). All species encountered were captured and identified. The salmonids were measured
(fork length) and weighed. Scales were removed from a representative sample and retained
for age analysis. After capture all fish were held in keep-nets to ensure their complete
recovery before being returned to the river.
14
Fig. 2: Illustration of the paired site selection approach in the King’s River catchment, Co.
Wicklow. Site DWW2 was paired with DWW15 and DWW5 was paired with DWW15.
• Control Sites
• Forested Sites
n = 19 paired sites
Fig. 3: Location of electrofishing site pairs
15
Table 4: Location of paired fishing sites
Main System
River
Site Code
Control/Forest
Easting
Northing
Main System
River
Site Code
Control/Forest
Easting
Northing
Kings
Ballinagee
BALLIN1
DWW1
Control
304462
204045
Kings
Annalecka
ANNA1
DWW13
25-50%
306426
202755
Kings
Glencreemore
GLEEN1
DWW5
Control
302788
200283
Kings
Glashaboy
GLASH1
DWW26
25-50%
306535
201611
Kings
Knickeen
KNICK1
DWW6
Control
299726
195072
Oilitigh
Oilitigh
OILI1
DWW17a
25-50%
299286
196067
Cloghoge
Cloghoge
CLOG1
DWW8
Control
312761
207455
Inchavore
Inchavore
INCH1
DWW19
25-50%
311004
206075
Srahmore
Srahmore
SRAH1a
DM11a
Control
096560
305240
Srahmore
Srahmore
SRAH2
DM23
>50%
095227
306980
Srahmore
Glenamong
GAMON4
DM8
Control
093918
304178
Srahmore
Glenamong
GAMON2
DM10
25-50%
092809
303819
Srahmore
Glenamong
GAMON5
DM9
Control
094080
303977
Srahmore
Glenamong
GAMON6
DM9a
25-50%
094019
303524
Owengarve
Callowswallagh
CALLOW1
DM22
Control
093465
298191
Owengarve
Glendahurk
CALLOW1
DM3
25-50%
091103
300931
Glenamoy
Glenamoy
GMOY3
DM19
Control
093801
332628
Glenamoy
Glenamoy
GMOY4
DM20/M8
25-50%
095106
335854
Glenamong
Fiddaunatoreen
FREEN1
DM6
Control
095099
301909
Glenamong
Glenamong
GAMON1
DM7
25-50%
094472
302777
Glenumerra/Glendavock
Glenumerra
GLENU1
DM24
Control
085739
267708
Glenumerra/Glendavock
Glenumerra
GLENU2
DM25
25-50%
089975
267651
Owenree
Owenree
OREE1
DG11
Control
101586
246870
Owenwee
Owenwee
OWEE2
DG22
>50%
103160
245498
Maumwee L. Inflow
Maumwee
MAUM1
DG24
Control
097255
248472
Owenwee
Owenwee
OWEE3
DG23
>50%
103292
245896
Owenriff
Owenriff
ORIFF1a
DG33
Control
105151
242453
Owenriff
Glengawbeg
GBEG1a
DG27
25-50%
106686
240525
Owenboliska
Owenboliska
OLISKA3
DG7
Control
111162
234916
Owenboliska
Owenboliska
OLISKA1
DG13
>50%
114582
235506
Owenboliska
Owenboliska
OLISKA3
DG7
Control
111162
234916
Owenboliska
Owenboliska
OLISKA6
DG15
>50%
108519
232725
Glenleheen
Glenleheen
GHEEN1
DD18
Control
190732
404350
Gweebara
Gweebara
GBAR3
DD11
25-50%
185974
402744
Deele
Deele
DEEL1
DD2
Control
211261
403208
Deele
Deele
DEEL2
DD19
5-25%
208901
405512
Elatagh
Elatagh
ELAT3
DD16
Control
202238
405218
Elatagh
Elatagh
ELATA5
DD14a
>50%
204295
403973
16
2.5 GEOGRAPHICAL IMAGING AND SITE CHARACTERISATION
Site co-ordinate readings were recorded from a GPS handset at all sites. Arcview™ 3.3 was
used to plot site distributions and delineated catchment basins for all sites. The
Geoprocessor™ extension program allowed the calculation of various catchment
characteristics including geology, soil (and sub-soil) coverage composition, percentage
catchment forestry, catchment land-use and catchment area. Catchment delineation was
performed by Compass Informatics™. Catchment characteristics were derived using this
delineation, the EPA geology database and the most up-to date FIPS database. FIPS 07 was
under development during this process and represented the best available data in March 2008.
The key information extracted from FIPS 07 included species composition, forest cover, and
felling history. The age of the tress was not available. Daniel McInerney, SBES, UCD,
undertook in part the geo-processing as outlined.
As mentioned previously two broad geology categories were assigned on the basis of the
dominant rock grouping, igneous/metamorphic or sedimentary. The igneous/metamorphic
category was predominantly composed of granite, but also included mica schist, quartzite,
Diorite, Gabbro and Dolerite while the sedimentary group included mainly Old Red
Sandstone, Coal Measures and some Carbonate geology.
Soils were assigned to four groups.
This categorisation followed an agreement on
interpretation of Irish Forest Soils and Teagasc Soil Categories. The initial catchment soil
mapping was based on the Teagasc Soils Map database distributed by the EPA. However,
the accuracy of this was unclear.
Subsequently, the Forest Service undertook to cross
reference the IFS site specific data from 1,732 sites (re-categorised as per the four soil
categories) with that held on the Teagasc National Soil cover data. The highest (73%)
agreement was within peats, followed by well drained minerals/gleys (62%).
Lower
percentage agreement was obtained for poorly drained gleys (46%), podzolic/lithosolics
(37%) and peaty gleys (21%).
Additional soil surveying in a selection of the study
catchments was deemed necessary to validate the soil designation. Priority was given to
catchments where sites showed variable responses in the aquatic biota to the presence of
forestry. A total of 106 catchments were targeted for soil surveying. Using a series of
systematic grids (250, 300 and 500m), 1,196 sample points were created and sampled. The
attributes data captured was based on the NFI methodology. The results again highlighted
17
the highest confidence in the designation of peats and well drained mineral soils (report by
John Redmond to WRBD). When the catchment soil allocations were finalised the dominant
soil type was used in all subsequent analyses as it was considered to have the greatest
potential for influencing the stream hydrochemistry.
2.6 STATISTICAL ANALYSES
Extensive databases for biological and physico-chemical parameters were generated in
Excel™. Univariate and multivariate analyses were performed using SPSS™ v. 12.0.1,
STATISITICA™ v. 7.1, Community Analysis Package (CAP™ v. 3.1) and Ecological
Community Analysis (ECOM™ v. 2.0). The AQEM Project (ASTERICS 3.10™) program
was used to generate over 40 water quality and macroinvertebrate metrics [using Europe
version].
Impairment in terms of the various biological metrics and hydrochemical
parameters was detected using metric values outside of two standard deviations (or 95%
confidence interval) of the control site values as expressed by Resh et al. (1988). A similar
approach was used by Johnson et al. (2005) to develop a clearfelling impact metric. Data
from 1 st and 2nd order sites were combined following preliminary analyses which indicated no
significant relationship between catchment size and taxon richness. These catchment sizes
ranged from 21.4 to 661.8ha. Sites with catchment sizes greater than 700ha were excluded
but were included in the analyses of longitudinal patterns. Sites with catchments less than
18ha were also excluded.
The hydrochemical data were used to derive means and minimum/maximum values for each
parameter. The minimum/maximum values were considered to represent the worst case
scenarios and were used to test relationships with forest cover, other catchment descriptors
and hydrochemical variables as well as the biological metrics.
Cluster analyses was carried out on the hydrochemical and biological datasets. Clustering is
the process of finding groups of objects (or data) such that those in a group are similar (or
related) to one another and different from (or unrelated to) the objects in other groups. Some
defined distance measure such as the Euclidean distance is often used to determine proximity
of the data in a cluster. The k-means clustering algorithm (Hartigan and Wong, 1979) is one
of the simplest unsupervised learning algorithms for this partitioning when the number of
18
clusters (k) is known or specified a priori. A good method will produce high quality clusters
with high intra-class similarity and low inter-class similarity (see figure below). The quality
of a clustering method is measured by its ability to discover some or all of the hidden
patterns. The quality of a clustering result also depends on both the similarity measure (like
Simpsons, Bray Curtis of Jaccards) used by the method and its implementation.
In regression analysis or modelling, the clustering helps determine if there are groups of
similar data that might exhibit a similar response ( which might require a specific model or
set of parameters) and also if the available data do not cover or span the region of interest.
When applied to response variables it clusters those with a similar pattern of responses
(which may or may not have a specific physical interpretation).
The k-means algorithm (Hartigan & Wong, 1979) used in the present study is one of the
simplest numerical methods used to implement clustering and works as follows:
1. The number of clusters required must be chosen in advance and a significance
tolerance for stopping the iterations.
2. An initial position in the data space is chosen for each cluster. These should be as far
apart as possible and should cover the range of the data space as well as possible.
3. The Euclidian distance from each point in the data set to all cluster centroids is
calculated and each data point is then associated with the nearest centroid. Thus a
cluster of data points is associated with each centriod.
4. The actual centroid of the points associated with each cluster is calculated and
replaces the previous centriod of that cluster.
19
5. Steps 3 and 4 are repeated until the change in the centroid positions is less than some
specified tolerance.
6. The solution is the set of clusters when the tolerance is satisfied.
Clustering was carried out separately using metrics to describe the chemical signature
(hydrogen (max), alkalinity (max) monomeric aluminium (max) cations (min), DOC
(max) and organic acids), physical descriptors that may affect the magnitude of the
pressure and run-off potential (% coniferous forest, catchment area, slope & area) and
selected
biological
response
metrics
(taxon
richness
&
ephemeroptera
richness/abundance).
2.7 QUALITY CONTROL
Quality control procedures were employed for macroinvertebrate sorting and identification.
Previously sorted samples were re-checked for missed specimens to check for % accuracy.
At most 10 individuals were recovered representing well below 3% of the total
macroinvertebrates initially sorted from the samples. A number of specimens from each
identified taxon were checked by an independent taxonomist, Dr Gustavo Becerra Jurado.
Quality control of data inputting to the physico-chemical and biological databases was also
undertaken. The macroinvertebrate databases from UCD and UCC were reviewed for
inconsistencies in taxonomy.
20
3. RESULTS
3.1 HYDROCHEMISTRY
Conductivity
The river sites examined were typically low conductivity waters. In fact, over 80% of the
sites recorded maximum conductivity reading below 150 S/cm (Fig. 4). The highest value
recorded was 295
S/cm in a tributary of the River Loobagh which drains sedimentary
geology. Overall, there was no significant differences in the mean readings across geology
and soil site groupings (Fig. 5) although the sedimentary sites on well drained mineral soils
had marginally higher values. In terms of a forest effect significant differences were detected
across the forest cover bands only in sites draining igneous/metamorphic geology and peaty
soils (Kruskall Wallis-H(3,65)=13.328, P=0.004).
Counts
100
100%
80
80%
60
60%
40
40%
20
20%
0
0%
0-50
50-100
100-150
150-200
200-250
250-300
Conductivity Range ( S/cm)
Fig. 4: Frequency distribution and accumulative percentage of conductivity readings from all
dates and sites.
The differences between low and high flow readings were highly variable and were typically
less than 100 S/cm but one sedimentary site on peat recorded a difference of 205 S/cm, the
highest value was associated with low flow. In other sites the highest conductivity readings
were associated with high flow.
21
Granite/Metamorphic
Sedimentary
Mean Conductivity ( S/cm)
160
140
120
100
80
60
40
20
0
Peat
Podzolic/Lithosolic
Peat
Podzolic/Lithosolic
Gleys
Well Drained
Mineral
Fig. 5: Variation in mean conductivity reading across geology and soil groups.
pH
The pH readings for the various sampling dates were highly variable (Figs. 6 and 7). Most of
the sites appeared to be episodically acidic. Some, especially those draining well drained
mineral soils, were more circum-neutral in character. Much of the variation within sites
could be related to differences in flow conditions, the low pH values were generally
associated with elevated flow. However, as previously mentioned it was difficult to know the
stage of the hydrograph represented and full flood conditions were not encountered at many
of the sites. It is therefore possible that the highest acidity levels were not captured by the
sampling programme.
8.5
8
7.5
Peat
Podzolic/Lithosolic
7
pH
6.5
6
5.5
5
4.5
4
3.5
DG20
DWW16
DM12
M3
DWW20
DWW22
DWW15
DWW21
DWW17
DWW26
DM3
M2
G16
DM10
DM11
G15
DM7
DD7
DWW10
DM5
DWW12
DM8
DM2
DM17
DM1
DG1
DG3
DG24
DWW7
DM4
DD10
DM6
G7
DG21
DG22
DG12
DD9
DG16
DG19
DG18
DD13
DG15
G9
DG14
DD5
G8
DG13
DD14
DG17
DG23
DD16
DM15
DWW19
DWW13
DD4
DD15
M8
G11
DG30
M11
DD8
DG6
DG7
DD6
DWW23
G6
DG8
DG31
DG29
DG25
G18
M6
DD2
M7
DG9
G5
DD3
DG28
G4
DM16
DWW6
DG11
DWW4
DWW2
DD11
DWW8
DWW9
Fig. 6: Distribution of pH readings from sites draining peat and podzolic/lithosolic soils on
granite/metamorphic geology. Sites within each soil group are ordered according to
increasing forest cover as indicated by the green arrow. The various colours represents the
three sampling dates.
22
8
7.5
7
6.5
pH
6
5.5
5
Gleys
Peat
Well-drained
Mineral
4.5
Podzolic/Lithosolic
4
3.5
Fig. 7: Distribution of pH readings from sites draining various soils on sedimentary geology.
Sites within each soil group are ordered according to increasing forest cover as indicated by
the green arrow. The various colours represents the three sampling dates.
Despite the uncertainty relating to flow conditions it should be noted that a good number of
control and forested sites were sampled in any one area under the same weather/flow
conditions. The randomised sampling should permit assessment of pH changes in relation to
forest cover. The data were initially analysed across the forestry bands. Minimum pH was
selected for analysis of the worst case condition.
On igneous/metamorphic geology
minimum pH was significantly different across the forest bands (Soil Type: Peats Minimum
pH:
Kruskall-Wallis-H(3,55) = 15.8426, p = 0.0012; Soil Type: Podzolic Lithosolic
Minimum pH: Kruskall-Wallis -H(3,31) = 9.228, p = 0.0264 – Fig. 7). Some of the lowest
values were associated with high forest cover, particularly on peats. The results were similar
when maximum hydrogen ion concentrations were analysed.
8.0
7.5
7.0
6.5
6.0
5.5
4.5
SoilTypes: Peats
Forested(>50%)
Forested(25-50%)
Forested(5-25%)
Control (<5%)
Forested(>50%)
Forested(25-50%)
3.5
Forested(5-25%)
4.0
Control (<5%)
Minimum pH
5.0
Median
25%-75%
Non-Outlier Range
Outliers
Extremes
SoilTypes: Podzolic Lithosolic
Fig. 8: Box plots of minimum pH values for sites within four forest cover bands draining
granite/metamorphic catchments with different dominant soil types.
23
A similar pattern was recorded on sedimentary geology but none of the differences was
statistically significant. Although the pH of sites on well drained mineral soils decreased
Forested(>50%)
Forested(25-50%)
Forested(5-25%)
Control (<5%)
Forested(>50%)
Forested(25-50%)
Poorly Drained Gleys
Forested(>50%)
Forested(25-50%)
Control (<5%)
Forested(>50%)
Forested(25-50%)
Podzolic Lithosolic
Forested(5-25%)
Forested(5-25%)
Forested(5-25%)
Peats
8.5
8.0
7.5
7.0
6.5
6.0
5.5
5.0
4.5
4.0
3.5
Control (<5%)
Minimum pH
8.5
8.0
7.5
7.0
6.5
6.0
5.5
5.0
4.5
4.0
3.5
Control (<5%)
across the forestry bands the streams remained circum-neutral (Fig. 9).
Median
25%-75%
Non-Outlier Range
Outliers
Extremes
Well Drained Mineral
So
Fig. 9: Box plots of minimum pH values for sites within four forest cover bands draining
sedimentary geology with different dominant soil types.
The relationship between minimum pH and % forest cover was also examined for both
geological settings. On igneous/metamorphic geology streams draining peat showed a
significant decrease (r = -0.6834, p = 0.0000002) in pH with increasing forest cover (Fig. 10).
Although control and sites with low forest cover had some pH reading as low as the more
heavily forested sites the latter sites had fewer readings in the circum-neutral range. The
minimum pH for both peat and podzolic/lithosolic sites fell below the lower limit of the
control sites when forest cover exceeds values in the region of 25-30%. The relationship on
peat on sedimentary geology was also significant (r = -0.2515, p = 0.0505) and largely similar
to that on igneous/metamorphic geology except that more readings were in the circum-neutral
24
range, except when forest cover exceeded 80% (Fig. 10). The podzolic/lithosolic sites did
not show a significant correlation between minimum pH and % forest cover. However, it
should be noted that the minimum pH values fell below the lower limit for the control sites
when forest cover exceeded 60% (Fig. 11).
8.0
7.5
7.0
Minimum pH
6.5
6.0
5.5
5.0
4.5
4.0
3.5
-20
0
20
40
60
80
100
120 -20
0
SoilTypes: Peats
20
40
60
80
100
120
SoilTypes: Podzolic Lithosolic
% Coniferous Cover
Fig. 10: Relationship between minimum pH and % forest cover for sites draining
granite/metamorphic geology with different dominant soil type.
Minimum pH
8.5
8.0
7.5
7.0
6.5
6.0
5.5
5.0
4.5
4.0
3.5
-20
0
20
40
60
80
100
120 -20
0
20
Peats
8.5
8.0
7.5
7.0
6.5
6.0
5.5
5.0
4.5
4.0
3.5
-20
0
20
40
40
60
80
100
120
80
100
120
Podzolic Lithosolic
60
80
100
120 -20
Poorly Drained Gleys
0
20
40
60
Well Drained Mineral
% Coniferous Cover
Fig. 11: Relationship between minimum pH and % forest cover for sites draining
sedimentary geology with different dominant soil types.
To further analyse these pH data for a possible forest effect it was hypothesised that the
number of pH readings below 5.0 would increase across the forest cover bands. This was
25
based on previous research in upland Wicklow streams which suggested that the duration of
low pH values in some forested streams exceeded that in control moorland streams (KellyQuinn et al., 1996a).
Tables 5 and 6 present the results for each geological setting. In
igneous/metamorphic catchments the number of pH readings <5.0 was substantially higher
than the result for the control sites in the 25-50% and >50% forestry bands for peat sites and
the 25-50% band for the podzolic/lithosolic sites.
Table 5: Numbers of pH readings 5 in each of the samplings rounds 1-3 and as %
percentage of overall samples for sites draining granite/metamorphic geology.
Soil Type
% Forest Cover
Band
1
2
3
Total
Samples
% total samples
with pH 5
<5
4/22
3/22
4/22
66
17
5-25.
1/9
1/8
0/7
24
8.3
25-50
4/6
2/7
3/8
21
43
>50
10/17
4/16
12/16
49
53
<5
0/11
2/8
1/8
11
7
5-25.
0/4
0/4
0/4
16
0
25-50
6/9
5/9
4/8
26
57
>50
2/8
0/8
1/8
24
8
Peat soil
Podzolic/Lithosolic
The result was similar for peat sites on sedimentary geology. On podzolic/lithosolic soils pH
values <5.0 were only encountered when forest cover exceeded 50%. The same applied to
sites on the poorly drained gleys. No pH readings <5 were recorded in catchments dominated
by well drained mineral soils (Table 6).
26
Table 6: Numbers of pH readings 5 in each of the samplings rounds 1-3 and as %
percentage of overall samples for sites draining sedimentary geology.
Soil Type
% Forest
Cover Bands
1
2
3
Total Samples
% total samples
with pH 5
<5
1/17
1/16
0/15
32
4
5-25.
1/4
1/4
0/4
12
17
25-50
2/10
2/10
2/9
29
21
>50
7/27
3/27
3/27
81
16
<5
0/13
0/13
0/13
39
0
5-25.
0/2
0/1
0/2
6
0
25-50
0/2
0/2
0/2
6
0
>50
0/22
3/22
4/22
66
11
<5
0/6
0/6
0/6
18
0
5-25.
0/3
0/3
0/3
9
0
25-50
0/7
0/7
0/5
19
0
>50
0/11
1/11
0/11
33
3
<5
0/2
0/4
0/2
6
0
5-25.
0/3
0/3
0/6
13
0
25-50
0/7
0/7
0/6
20
0
>50
0/9
0/9
0/9
27
0
Peat
Podzolic/Lithosolic
Poorly drained Gleys
Well-drained Mineral
70
60
Peat
Podzolic/Lithosolic
% Records.
50
40
30
20
10
0
2
4
6
8
10
12
14
16
18
20
>20
Alkalinity (mg/l CaCO3)
Fig. 12: Distribution of alkalinity readings from sites on predominately peat and
podzolic/lithosolic soils in igneous/metamorphic catchments.
27
Alkalinity, Sodium Dominance Index and Aluminium
The majority of the alkalinity readings from sites draining igneous/metamorphic geology
with either peat or podzolic/lithosolic soil cover fell below 2 mg/l CaCo3 (Fig. 12). A few
sites had values >20 mg/l CaCO3, these catchments were influenced by variable amounts of
carbonate geology in the catchment.
On the same soils in catchments dominated by
sedimentary geology readings were more evenly distributed across the alkalinity range with
over 20% higher than 20 mg/l CaCO3 (Fig. 13).
% Records.
70
60
Peat
50
Podzolic/Lithosolic
40
30
20
10
0
2
4
6
8
10
12
14
16
18
20
>20
Alkalinity (mg/l CaCO3)
Fig.13: Distribution of alkalinity readings from sites on predominately peat and
podzolic/lithosolic soils on sedimentary geology.
As expected, sites draining well drained mineral soil were more buffered and the majority of
readings were greater than 20 mg/l CaCO3 (Fig. 14).
70
60
Gleys
% Records.
50
Well Drained Mineral
40
30
20
10
0
2
4
6
8
10
12
14
16
18
20
>20
Alkalinity (mg/l CaCO3)
Fig. 14: Distribution of alkalinity readings for sites on predominately peat and
podzolic/lithosolic soils in igneous/metamorphic catchments.
28
In terms of minimum alkalinity the highest number of values <8mg/l CaCO3 were associated
with
the
following
geological
setting;
podzolic/lithosolic/igneous/metamorphic
peat/igneous/metamorphic
(95%),
peat/sedimentary
(96%),
(72%),
podzolic/lithosolic/sedimentary (52%). The presence of forestry tended to depress the site
alkalinity as can be seen from Fig.15 which compares the distribution of alkalinity readings
from control and heavily forested sites on peat and igneous geology. The effects of forestry
were most obvious when minimum alkalinity was considered. Minimum alkalinity decreased
significantly across the forest cover bands on peat/igneous/metamorphic (Kruskall-WallisH(3,55) = 14.6122, p = 0.0022), podzolic/lithosolic/ igneous/metamorphic (Kruskall-Wallis H(3,31) = 8.0601, p = 0.0448), but this trend was not statistically significant in the other
geological settings. However, in the case of peat on granite the occurrence of negative
minimum alkalinity increased across the forest cover range. Only the forested sites on
podzolic/lithosolic soils recorded negative alkalinity values, some were detected on
igneous/metamorphic geology when forest cover exceeded circa 25%.
On sedimentary
geology negative alkalinity values were recorded at some sites when forest cover exceeded
% Records.
60%. It should be noted that not all forested sites exhibited this loss of buffering capacity.
50
45
40
35
30
25
20
15
10
5
0
Peat on Granite/Metamorphic - <5% Forest Cover
Peat on Granite/Metamorphic - >50% Forest Cover
0
2
4
6
8
10
12
14
16
Alkalinity (mg/l CaCO3)
Fig. 15: Distribution of alkalinity readings for sites on predominately peat soils in
igneous/metamorphic catchments with and without forest cover.
The relationship between maximum and minimum alkalinity values illustrates the level of
change in buffering capacity between the two extremes. The majority of the sites that
recorded zero or negative alkalinity had maximum alkalinity below 8 mg/l CaCO3 (Fig. 16a).
29
In contrast, several forested sites with maximum alkalinity up to 20 mg/l CaCO3 and higher
exhibit zero or negative minimum alkalinity values (Fig. 16b).
70
60
50
2
R = 0.6422
40
30
20
.
Min. Alkalinity (mg/l CaCO3)
80
10
0
-20
-10
0
20
40
60
80
100
120
140
160
Max. Alkalinity (mg/l CaCO3)
(a)
80
Min. Alkalinity (mg/l CaCO3)
70
2
60
R = 0.6893
50
40
30
20
10
0
-20
-10
0
20
40
60
80
100
120
140
160
Max. Alkalinity (mg/l CaCO3)
(b)
Fig. 16: Relationship between maximum and minimum alkalinity for (a) control sites,(b) sites
with >25% forest cover.
Over 90% of the sites dominated by peat or podzolic/lithosolic soil on igneous/metamorphic
geology had Sodium Dominance values >60% supporting their acid-sensitive designation.
On sedimentary geology the results were more variable, with an increasing proportion of the
sites falling below 50% SDI as one moved from peat, through podzolic/lithoslic soils to the
gleys and well drained mineral soils. In addition, there were greater differences between
maximum and minimum SDI values.
30
Total aluminium concentrations were highest at sites draining predominantly peat (Figs. 17 &
18). In the two geological settings total aluminium increased significantly with increasing
forest cover. The trend for sites influenced by podzolic/lithosolic soils was only significant
on sedimentary geology.
600
500
Maximum Total Aluminium
µg/l)
(
400
300
200
100
0
-100
-20
0
20
40
60
80
100
120 -20
0
SoilTypes: Peats
20
40
60
80
100
120
SoilTypes: Podzolic Lithosolic
% Coniferous Cover
Fig. 17: Relationship between forest cover and maximum total aluminium concentrations on
Maximum Total Aluminium (µg/l)
granite/metamorphic geology with peat and podzolic/lithosolic soils.
800
700
600
500
400
300
200
100
0
-100
-20
0
20
40
60
80
100
120 -20
0
20
Peats
800
700
600
500
400
300
200
100
0
-100
-20
0
20
40
40
60
80
100
120
80
100
120
Podzolic Lithosolic
60
80
100
120 -20
Poorly Drained Gleys
0
20
40
60
Well Drained Mineral
% Coniferous Cover
Fig. 18: Relationship between forest cover and maximum total aluminium concentrations on
sedimentary geology with various dominant soil types.
31
Few measurement of labile monomeric aluminium were made and most were for forested
sites. Maximum aluminium concentrations ranged from 17.0 – 348 ug/l. No significant
correlation with forest cover was detected which may be a factor of the sample size.
Sources of Acidity
Dissolved Total Organic Carbon (DTOC)
As expected the variation in background DTOC concentrations, as illustrated by the control
sites, reflected the organic nature of the dominant catchment soils.
The highest
concentrations were recorded from sites draining peat on igneous/metamorphic geology (Fig.
19) followed by peat on sedimentary geology (Fig. 20).
In these two settings DTOC
concentrations showed a significant increase across the forest cover bands (Peats on
Igneous/metamorphic-Kruskall Wallis-H(3,55) = 19.7422, P = 0.0002; Peats on SedimentaryKruskall Wallis-H(3,61) = 12.5833, p = 0.0056).
.
45
40
35
Maximum DOC (mg/l)
30
25
20
15
10
5
0
-5
Control (<5%)
Forested(25-50%)
Control (<5%)
Forested(25-50%)
Fores ted(5-25%)
Forested(>50%)
Forested(5-25%)
Fores ted(>50%)
SoilTypes: Peats
Median
25%-75%
Non-Outlier Range
Outliers
Extremes
SoilTypes: Podzolic Lithosolic
Fig. 19: Box plots of maximum DTOC values from sites within four forest cover bands
draining igneous/metamorphic catchments with different dominant soil types.
32
60
50
40
30
20
10
Maximum DOC (mg/l)
0
-10
Control (<5%)
Forested(25-50%)
Forested(5-25%)
Forested(>50%)
Control (<5%)
Forested(25-50%)
Forested(5-25%)
Forested(>50%)
Peats
Podzolic Lithosolic
60
50
40
30
20
10
0
-10
Control (<5%)
Forested(25-50%)
Forested(5-25%)
Forested(>50%)
Control (<5%)
Forested(25-50%)
Forested(5-25%)
Forested(>50%)
Poorly Drained Gleys
Well Drained Mineral
Median
25%-75%
Non-Outlier Range
Outliers
Extremes
Fig. 20: Box plots of maximum DTOC values from sites within four forest cover bands
draining sedimentary geology with different dominant soil types.
Excess or Non-Marine Sulphate
Maximum non-marine sulphate differed significantly across forestry bands for sites draining
peat in both geological settings (Igneous/metamorphic – Kruskall-Wallis-H(3,55) = 8.1725, p
= 0.0426; Sedimentary Geology – Kruskall-Wallis-H(3,62) = 25.0755, p = 0.00001) The
trend was similar for podzolic/lithosolic soils but was only significant on sedimentary
geology. Interestingly on mineral soils non-marine sulphate decreased significantly across
the forest-cover bands (Kruskall-Wallis-H(3,34) = 9.2423, p = 0.0262).
Nitrate
Sites draining peats and podzolic/lithosolic soils in both geological settings recorded
maximum nitrate concentrations largely below 0.5 mg/l NO3 and there was no significant
correlation with forest cover. Concentrations were higher at sites on gleys (0.05-2.75 mg/l
NO3) and well drained mineral soils (0.18-6.22 mg/l NO3). The latter sites recorded a
significant decrease in nitrate concentration across the forest cover gradient.
33
Chloride
Maximum chloride concentrations ranged from 6.10 to 44.96 mg/l at sites draining peat on
igneous/metamorphic geology. Values were up to 10mg/l lower at peat sites on sedimentary
geology and marginally lower at sites draining podzolic/lithosolic soils. The relationship
with forest cover was only significant for sites on peat in both geological settings.
Calculations suggested that dilution makes a variable contribution to loss of alkalinity and in
many cases the forested sites showed a slightly higher % value. Anion titration was detected
in all events examined. The principal contributors were organic acids and sulphate. Excess
sulphate only made a contribution in the Wicklow sites and at one site in Galway. The
contribution of nitrate across all sites was insignificant. The contribution of sea salts to
acidification was also low and only one significant sea-salt event was detected at one site in
Galway.
Comparison of Source Streams – Upstream and Downstream of Forestry
On igneous geology, no significant differences for any of the chemical variables were found
between control sources and the sources sampled above forests (both 5-25% and >25% forest
bands), (Mann-Whitney, P>0.05).
A similar result was noted on sedimentary geology
(Mann-Whitney, P>0.05). However, it should be noted that on sedimentary geology, only
two sites were sampled above the forestry in the 5-25% and >25% categories. Therefore,
both of these forestry bands had to be combined into a single forest category and compared to
the control sources. This result suggests that all sources (control and above forests) had no
significant differences.
Control sources did not differ significantly from the downstream sites on the same stream in
terms of pH and alkalinity (Wilcoxon Ranked Sign Test, P>0.05). However, chloride, NM
sulphate and sodium were significantly higher (Wilcoxon; Chloride: Z = -2.757, P = 0.006;
NM Sulphate: Z = -2.114, P = 0.034; Sodium: Z = -3.371, P = 0.001) downstream. Sites
downstream of the 5-25% forested band on igneous/metamorphic geology differed
significantly from their sources in terms of pH, SDI, chloride, sulphate, NM sulphate,
sodium, magnesium, calcium, NM calcium, total hardness and non-marine hardness
(Wilcoxon, P<0.05).
Sites downstream of >25% forest cover had significantly higher total
34
monomeric aluminium, chloride, sulphate, NM sulphate and sodium (Wilcoxon, P<0.05) than
their respective sources.
On sedimentary geology sites, the downstream control sites recorded significant differences
from their corresponding sources for chloride, sulphate, NM sulphate, sodium, NM sodium
and NM magnesium (Wilcoxon, P<0.05). On sedimentary geology only one site pairing
represented the 5-25% forest cover category however, a difference in NM Ca was noted
(Wilcoxon, P<0.01). The >25% forested category on sedimentary geology presented
significant differences between downstream and source for pH, hydrogen, alkalinity, SDI,
NM sodium, magnesium, NM magnesium, calcium, NM calcium, total hardness and NM
hardness (Wilcoxon, P<0.05 and P<0.001). Results are presented in Table 7. The higher
sodium levels at downstream sites were accompanied by higher magnesium and calcium
values. This maintained the SDI values as the overall ratio of cations remained quite similar
at the source and downstream sites.
Table 7: Significant results from paired analysis for selected chemical variables at
downstream and source sites (Wilcoxon Ranked Sign Test).
Igneous
Metamorphic Sites
Control <5% Forest
5-15% Forest Cover
>25% Forest cover
Parameter
Chloride
NM Sulphate
Sodium
Wilcoxon
(Z)
2.757
-2.114
-2.371
P
value
0.006
0.034
0.001
pH
Chloride
-1.503
-3.11
0.028
0.002
SDI
Sulphate
NM Sulphate
Sodium
Magnesium
Calcium
NM Calcium
Total Hardness
-2.062
-2.97
-3.11
-3.18
-2.551
-2.831
-2.481
-2.9
0.039
0.003
0.001
0.001
0.011
0.005
0.013
0.004
Total
Aluminium
Chloride
Sulphate
-3.068
0.002
-2.425
-2.516
0.001
0.012
NM Sulphate
Sodium
-2.0
-3.555
0.012
<0.001
Sedimentary Sites
Parameter
Wilcoxon
(Z)
-3.516
-3.206
-2.999
-2.999
-3.154
-2.223
P
value
<0.001
0.001
0.003
0.003
0.002
0.026
Control <5% Forest
Chloride
Sulphate
NM Sulphate
Sodium
NM Sodium
NM
Magnesium
5-15% Forest Cover
NM Calcium
-2.201
0.028
>25% Forest cover
pH
Hydrogen
Alkalinity
SDI
NM Sodium
Magnesium
NM
Magnesium
Calcium
-2.971
-2.621
-2.345
-2.342
-2.201
-2.622
-2.271
0.003
0.009
0.019
0.019
0.028
0.009
0.023
-2.622
0.009
NM Calcium
Total
Hardness
NM Hardness
-2.411
-2.691
0.016
0.007
-2.621
0.009
Although the above analyses highlighted only two site grouping that recorded lower
downstream pH than at the sources there were several individual sites within other groups
35
that followed this pattern. Several of the igneous sites in Co. Wicklow were more acidic
downstream than their corresponding sources. These sites included those on the Annalecka,
Lugduff and Glashaboy rivers. These sites had ~40-70% catchment cover of coniferous
forest.
36
3.2 MACROINVERTEBRATES
Community Composition
In total, over 318,000 individual specimens were sorted and identified to the lowest possible
taxonomic level from the 239 study sites. These yielded a total of 204 distinct taxa. The
most diverse group was the Trichoptera followed by the Coleoptera (Table 8).
Table 8: Taxon richness in the major taxonomic groups
Taxon
Trichoptera
Coleoptera
Diptera
Ephemeroptera
Plecoptera
Gastropoda
Odonata
Crustacea
Hirudinea
Hemiptera
Neuroptera
Lamellibranchia
Richness
61
53
29
18
17
8
5
3
4
2
1
1
Some of these taxa were highly localised, such as the mayfly species, Ameletus inopinatus
Eaton, found only in samples collected in Wicklow and Donegal. Other mayfly, such as
Baetis rhodani (Pictet.) and Leptophlebia vespertina (Linn.) were more ubiquitous. Species
such as Caenis rivulorum Eaton and the caddis-fly, Sericostoma personatum (Kirby &
Spence) were considered acid-sensitive as they were located in more buffered regions on
sedimentary geology.
More acid-tolerant species including, Ameletus inopinatus,
Siphlonurus lacustris (Eaton), Leptopheblia vespertina and Plectrocnemia conspersa (Curtis)
were present in higher abundances in areas of weathering tolerant, acid-sensitive, igneous
geologies.
37
In general the mean abundances of macroinvertebrates was significantly higher at the
sedimentary sites (One-way ANOVA; F(1,5) = 59.058, P = 0.002; Fig. 21).
700
Sedimentary
Igneous/Metamorphic
600
Abundance
500
400
300
200
100
0
Control
5-25% Forest
Cover
25-50% Forest
Cover
>50% Forest
Cover
Fig. 21: Mean macroinvertebrate abundances at the sedimentary and igneous/metamorphic
sites across the four forest cover bands.
The higher total macroinvertebrate abundances at the sedimentary sites could be largely
attributed to the Ephemeroptera and Chironomidae (Fig. 22). In both geological settings the
Ephemeroptera was reduced in abundance at sites in the two highest forest cover bands. At
the igneous/metamorphic sites the reduction in ephemeropteran abundance was largely
balanced by an increase in the numbers of Plecoptera. This did not occur at the sites draining
sedimentary geology and consequently overall abundance declined gradually across the forest
cover bands.
Selection of Macroinvertebrates Metrics
Approximately 45 different water quality and diversity metrics were generated for the dataset
using the AQEM (ASTERICS 3.10™) computer software. Those which were most
appropriate for Ireland and which showed a significant correlation with pH were selected to
detect impacts due to acidification. These included taxon richness (Fig. 23), ephemeropteran
richness, ephemeropteran abundance, trichopteran richness, Baetis abundance, %EPT.
38
250
Control
5-25% Forest Cover
25-50% Forest Cover
>50% Forest Cover
Abundance
200
150
100
50
Ot
he
rs
Ch
iro
no
m
id
ae
Si
m
ul
iid
ae
Cr
us
ta
ce
a
Tr
ich
op
te
ra
Pl
ec
op
te
ra
Ep
he
m
er
op
te
ra
0
(a)
250
Control
5-25% Forest Cover
25-50% Forest Cover
>50% Forest Cover
Abundance
200
150
100
50
Ot
he
rs
Si
m
ul
iid
ae
Ch
iro
no
m
id
ae
Cr
us
tac
ea
Tr
ich
op
ter
a
Pl
ec
op
te
ra
Ep
he
m
er
op
te
ra
0
(b)
Fig. 22: Mean abundances of the major taxonomic groups at sites on (a)
igneous/metamorphic and (b) sedimentary geology. Standard error bars are included.
pH(min):Taxon Richness: r2 = 0.2437; r = 0.4937, p = 0.0000
60
Taxon Richness
50
40
30
20
10
0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
pH (minimum)
Fig. 23: Relationship between taxon richness and minimum pH across all sites.
39
Taxon Richness
On igneous/metamorphic geology taxon richness ranged from 11 to 52, the lower value was
from a forested catchment (DWW22, 67% forest cover). The range of values was similar on
sedimentary geology (16 at L5, 37% forest cover to 55 at one of the control sites). There was
a significant decline in taxon richness with increasing coniferous forest cover on peat and
well drained mineral sites on sedimentary geology (Kruskall-Wallace; Sedimentary/Peat –
H(2,65) = 10.4252, P<0.05; Sedimentary/Well Drained Mineral – H(2,34) = 9,4919, P<0.05, Fig.
24). The differences were not statistically significant for igneous/metamorphic sites.
However, the number of sites with taxon richness less than 30 was highest in the >50% forest
cover band compared to the control group (Fig. 25) draining peats, and the pattern was
retained when sites on podzolic/lithosolic soils were added to the analysis (Fig. 26).
Taxon Richness
60
55
50
45
40
35
30
25
20
15
10
Control(<5%)
Forested(25-50%)
Control(<5%)
Forested(25-50%)
Forested(5-25%)
Forested(>50%)
Forested(5-25%)
Forested(>50%)
Peats
Podzolic Lithosolic
60
55
50
45
40
35
30
25
20
15
10
Control(<5%)
Forested(25-50%)
Control(<5%)
Forested(25-50%)
Forested(5-25%)
Forested(>50%)
Forested(5-25%)
Forested(>50%)
Poorly Drained Gleys
Median
25%-75%
Non-Outlier Ran
Outliers
Extremes
Well Drained Mineral
Fig. 24: Box plots of taxon richness for sites on sedimentary geology with different soil
categories.
40
Control
>50% Forest Cover
...............
% Sites
45
40
35
30
25
20
15
10
5
0
20
25
30
Taxon Richness
>30
.
Fig. 25: Comparison of the distribution of taxon richness at sites draining peat on
igneous/metamorphic geology.
70
60
50
>50% Forest Cover
25-50% Forest Cover
5-25% Forest Cover
Control
% Sites
40
30
20
10
0
10
15
20
25
30
>30
Taxon Richness
Fig. 26: Distribution of taxon richness counts at sites draining peat and podzolic/lithosolic
soils on igneous/metamorphic geology.
Ephemeropteran Richness
Ephemeropteran richness reached a maximum of 8 species on igneous/metamorphic geology
with one additional species on the sedimentary geology. In both settings some of the forested
sites recorded a low diversity of Ephemeroptera.
Ephemeropteran richness was significantly negatively correlated with % forest cover at sites
draining peats (r=-0.4640, P<0.001) and podzolic lithosols (r=-0.3884, P<0.05) on
igneous/metamorphic geology. As forest cover increased an increasing number of sites
41
recorded low taxon richness (Fig. 27). On sedimentary geology (Fig. 28) ephemeropteran
richness was again significantly correlated to percentage conifer cover on peat (r = -0.5378,
P<0.001) and well drained mineral soils (r = -0.4855, P<0.001).
9
8
Ephemeropteran Richness
7
6
5
4
3
2
1
0
-1
-20
0
20
40
60
80
100
120 -20
0
20
SoilTypes: Peats
40
60
80
100
120
SoilTypes: Podzolic Lithosolic
% Coniferous Cover
Fig. 27: Relationship between ephemeropteran richness and forest cover at sites draining peat
and podzolic/lithosolic soils on igneous/metamorphic geology.
10
8
6
4
Ephemeropteran Richness
2
0
-2
-20
0
20
40
60
80
100
120 -20
0
20
Peats
40
60
80
100
120
80
100
120
Podzolic Lithosolic
10
8
6
4
2
0
-2
-20
0
20
40
60
80
100
120 -20
0
Poorly Drained Gleys
20
40
60
Well Drained Mineral
% Coniferous Cover
Fig. 28: Relationship between ephemeropteran richness and forest cover at sites draining
different soils on sedimentary geology.
42
As was highlighted for total taxon richness, an increasing number of sites showed a reduction
in ephemeropteran richness (Table 9) along the forest cover gradient. Some 13% of sites
draining peat on igneous/metamorphic geology in the >50% coniferous cover band were
devoid of Ephemeroptera, while a further 75% only had one species present.
On
podzolic/lithosilic soils there relatively few Ephemeroptera at sites in the three forestry
bands. A decline in the occurrence of Baetis spp. occurred across the forestry bands. For
example 9% of the control sites on peat had low numbers of Baetis spp. compared to 71% of
sites in the >50% forest cover band. The replacement of Baetis by more acid-tolerant species
(e.g. Siphlonurus lacustris) was a feature of the latter group of sites.
Table 9: Percentage distribution of ephemeropteran taxon richness counts for sites on (a)
peats and (b) podzolic/lithosolic soils on igneous/metamorphic geology
(a)
Richness
0
1
2
3
4
5
6
7
8
(b)
Control
0
5
18
36
9
23
5
0
4
Forest
5-25%
0
22
33
11.5
0
11.5
22
0
0
Cover
25-50%
0
14
14.5
14
14.5
29
14
0
0
>50%
13
75
0
6
0
6
0
0
0
Richness
0
1
2
3
4
5
6
7
8
Control
0
18
9
9.5
36
9.5
9
0
9
Forest
5-25%
25
0
0
25
25
0
25
0
0
Cover
25-50%
11
44
22
11.5
0
0
0
11.5
0
>50%
25
25
12.5
12.5
25
0
0
0
0
The pattern was similar for peat sites on sedimentary geology where a low number of
Ephemeroptera was recorded in the two top forestry cover bands (Table 10).
On
podzolic/lithosolic soils only sites in the >50% band recorded no mayfly and a large
proportion of the sites in the >50% band had just one or two species present.
Baetis spp.
were only absent from sites (10%) on podzolic/lithosolic soils in the >50% forest cover band.
At sites on poorly drained gleys and well drained mineral soils there was little differences in
the distribution of ephemeropteran counts across the forest cover bands (Table 11).
43
Table 10: Percentage distribution of ephemeropteran taxon richness counts for sites draining
(a) peats and (b) podzolic/lithosolic soils on sedimentary geology.
(a)
Richness
0
1
2
3
4
5
6
7
8
9
(b)
Control
0
0
0
15
10
40
15
10
5
5
Forest
5-25%
0
50
0
25
0
0
0
25
0
0
Cover
25-50%
18.5
9
9
18.5
9
9
9
9
9
0
>50%
7
43
20
3.5
13
10
0
3.5
0
0
Richness
0
1
2
3
4
5
6
7
8
9
Control
0
0
0
0
23
23
31
15
8
0
Forest
5-25%
0
0
0
0
0
50
50
0
0
0
Cover
25-50%
0
0
0
0
50
0
0
0
50
0
>50%
8
8
8
8
17.5
25
17.5
8
0
0
Table 11: Percentage distribution of ephemeropteran taxon richness counts for sites on (a)
poorly drained gleys and (b) well drained mineral soils on sedimentary geology.
(a)
(b)
Richness
0
1
2
3
4
5
6
7
8
9
Control
0
33.3
0
33.3
33.3
0
0
0
0
0
Forest
5-25%
0
0
0
33.3
0
33.3
33.3
0
0
0
Cover
25-50%
0
0
14.5
14
14.5
14
29
14
0
0
>50%
0
33
0
0
22.5
22
22.5
0
0
0
Richness
0
1
2
3
4
5
6
7
8
9
Control
0
0
0
0
0
50
50
0
0
0
Forest
5-25%
0
0
20
0
40
20
0
20
0
0
Cover
25-50%
0
0
20
0
40
20
0
20
0
0
>50%
0
0
44.5
0
0
44.5
11
0
0
0
Ephemeropteran Abundance
A reduction in abundance of indicator taxa can often highlight environmental stress and it is
considered to be a useful early warning indicator of impact. No significant correlation was
detected between ephemeropteran abundance and forest cover for sites draining either peat or
podzolic/lithosolic soil types on igneous/metamorphic geology. The same applied to these
soil types on sedimentary geology. However, a significant relationship was detected for sites
located on well drained mineral soils on sedimentary geology (r = -0.6358, P>0.001).
44
700
600
500
400
300
200
Ephemeropteran Abundance
100
0
-100
-20
0
20
40
60
80
100
120 -20
0
20
Peats
40
60
80
100
120
80
100
120
Podzolic Lithosolic
700
600
500
400
300
200
100
0
-100
-20
0
20
40
60
80
100
120 -20
Poorly Drained Gleys
0
20
40
60
Well Drained Mineral
% Coniferous Cover
Fig. 29: Relationship between % forest cover and ephemeropteran abundance on sedimentary geology
The cluster of high mayfly abundance noted on sedimentary peats between pH levels 6.5 and
7.5 (Fig. 29) corresponds to a cluster of highly buffered, high pH sites in Co. Cork. Despite
high levels of coniferous forest cover at these sites, the occurrence of variable amounts of
mineral soils among the peats improved buffering capacity and allowed for higher
abundances of mayfly (particularly Baetis rhodani). Despite the lack of a strong correlations
between abundance and % forest cover it was clear that on peat and podzolic soil in both
geological settings the number of sites with low numbers (zero and <5) of ephemeropteran
specimens increased across the forest cover bands. The results are illustrated for sites on (a)
peat and (b) podzolic/lithosolic soils in Fig. 30.
45
(a)
70
Control
60
5-25% Forest Cover
% Sites
50
25-50% Forest Cover
40
>50% Forest Cover
30
20
10
0
0
5
10
15
20
>20
Ephemeropteran Abundance
% Sites
.
70
(b)
60
Control
50
5-25% Forest Cover
40
25-50% Forest Cover
>50% Forest Cover
30
20
10
0
0
5
10
15
20
>20
Ephemeropteran Abundance
Fig. 30: Frequency distribution of ephemeropteran abundance counts assigned to 6
abundance categories (0; ≤5; ≤10; ≤15; ≤20; >20) at sites draining (a) peat and (b)
podzolic/lithosolic soils on igneous/metamorphic geology.
EPT Metric
Ephemeropteran (E), plecopteran (P) and trichopteran (T) richness values are used to
calculate EP and EPT metric. In the present study EPT richness correlated significantly (R2=
0.735, P<0.0001) with EP richness and therefore only one of these was applied in the
analyses. The variability plot of EPT richness indicated a shift in distribution towards the
lower end of the scale as one moved across the forest cover bands (Fig. 31). While no
significant differences in median EPT was detected across forestry bands on
igneous/metamorphic geology (P>0.05), significant declines in both metrics were found for
peat sites on sedimentary geology (Kruskall-Wallis- EPT – H(3,65) = 10.0914, P = 0.0178, Fig.
32). A similar trend occurred on well drained mineral soil but the relationship was not
significant (P>0.05).
It is worth noting that the median EPT of sites draining
podzolic/lithosolic soils fell well below the control median in the >50% forest cover band.
46
35
EPT (Richness)
30
25
20
15
10
Sedimentary
Forested(>50%)
Forested(25-50%)
Control(<5%)
Forested(5-25%)
Forested(>50%)
WellDrainedMineral
Poorly Drained Gleys
Forested(25-50%)
Control(<5%)
Forested(5-25%)
Forested(>50%)
Forested(25-50%)
Control(<5%)
Forested(5-25%)
Podzolic Lithosolic
Forested(>50%)
Forested(25-50%)
Peats
Control(<5%)
Igneous-metamorphic
Forested(5-25%)
Forested(>50%)
Control(<5%)
WellDrainedMineral
Forested(25-50%)
Podzolic Lithosolic
Control(<5%)
Forested(5-25%)
Forested(>50%)
Forested(25-50%)
Peats
Control(<5%)
0
Forested(5-25%)
5
Control
SoilTypes
Geology
Fig. 31: Variability plot of EPT across coniferous forest bands for each geology and soil
category.
35
30
25
20
15
10
5
EPT (Richness)
0
Control(<5%)
Forested(25-50%)
Control(<5%)
Forested(25-50%)
Forested(5-25%)
Forested(>50%)
Forested(5-25%)
Forested(>50%)
Peats
Podzolic Lithosolic
35
30
25
20
15
10
5
0
Control(<5%)
Forested(25-50%)
Control(<5%)
Forested(25-50%)
Forested(5-25%)
Forested(>50%)
Forested(5-25%)
Forested(>50%)
Poorly Drained Gleys
Median
25%-75%
Non-Outlier Range
Outliers
Extremes
Well Drained Mineral
Fig. 32: Box plots for EPT richness on sites draining sedimentary geology.
47
The relationship
between EPT richness and % forest cover is further explored in the
correlation plots (Fig. 33). The correlation was significant for peat sites in both geological
SoilTypes: Podzolic
Lithosolic
SoilTypes: Poorly
Drained Gleys
EPT (Richness)
35
30
25
20
15
10
5
0
SoilTypes:
WellDrainedMineral
35
30
25
20
15
10
5
0
SoilTypes: Peats
Scatterplot (MetricsUCDAnalysis_29_4_08_JRB 241v*239c)
35
30
25
20
15
10
5
0
35
30
25
20
15
10
5
0
-20
0
20
40
60
80
100
120 -20
Geology: Sedimentary
0
20
40
60
80
100
120
Geology: Igneous-metamorphic
%Coniferous
Fig. 33: Relationship between EPT richness and % forest cover in the various geological
settings.
settings and site EPT richness began to fall below the lower limit of the control sites when
forest cover exceeded circa 25-30% . This also applies to sites on podzolic/lithosolic soils on
igneous/metamorphic geology. On sedimentary geology the podzolic sites recorded low EPT
above 50% forest cover.
BMWP & ASTP Metrics
While Kruskall-Wallis tests on BMWP did not show a significant difference between forest
bands on peats and podzolic/lithpsolic sites on igneous/metamorpgic geology there were
nonetheless strong trends of decreasing BMWP across the forest cover bands. Peat sites on
sedimentary geology demonstrated a significant decrease in BMWP with increasing forest
cover bands (H(3,65) = 10.3406, P = 0.0159). While the trend was only statistically significant
48
for peats some sites on podzolic/lithosols and poorly drained gleys showed a distinct decrease
in BMWP scores in the >50% forest cover band. The Biological Monitoring Working Party
(BMWP) Score was significantly correlated with % coniferous cover for sites draining peat
soils on both igneous/metamorphic (r = -0.3209, P = 0.03, Fig. 34 ) and sedimentary rock (r =
-0.4195, P = 0.0009) geology (Fig. 35).
The relationship was not significant for
podzolic/lithosolic soils but here again it should be noted that scores for sites on
igneous/metamorphic and sedimentary geology began to fall below the control BMWP scores
when forest cover exceed 25% and 60%, respectively. Average Score Per Taxa (ASTP)
significantly correlated with increasing % coniferous cover at sedimentary peats sites only (r
= -0.3679, P = 0.0041). A significant difference was also detected between forest cover
bands for sedimentary peat sites, with ASTP decreasing as forest bands increased (H(3.65) =
7.9399, P = 0.0473)
200
180
160
BMWP Score
140
120
100
80
60
40
20
-20
0
20
40
60
80
100
120 -20
SoilTypes: Peats
0
20
40
60
80
100
120
SoilTypes: Podzolic Lithosolic
BMWP Score
Fig. 34: Relationship between BMWP and % coniferous cover for soils on igneous/metamorphic geology.
220
200
180
160
140
120
100
80
60
40
20
-20
0
20
40
60
80
100
120 -20
0
20
Peats
220
200
180
160
140
120
100
80
60
40
20
-20
0
20
40
40
60
80
100
120
80
100
120
Podzolic Lithosolic
60
Poorly Drained Gleys
80
100
120 -20
0
20
40
60
Well Drained Mineral
Fig. 35: Relationship between BMWP scores and % forest cover for sites draining sedimentary
geology.
49
Community Diversity and Species Evenness
A trend of decreasing scores for both the Simpson and Margaley diversity indices was noted
across the forest cover bands. However, the differences between bands was only significant
for peat sites in both geological settings. The correlation with % forest cover was highly
significant (P=0.001) for sites draining peat on sedimentary geology (Fig. 36). Species
evenness varied greatly across sites and settings and none of the trends was statistically
significant at P<0.05. However, the pattern, previously discussed, whereby some forested
sites fell below the minimum values of the control sites at forest cover was repeated but the
number of sites involved was much fewer.
10
9
8
7
6
5
Diversity (Margalef Index
4
3
2
Control(<5%)
Forested(25-50%)
Fores ted(5-25%)
Forested(>50%)
Control(<5%)
Forested(25-50%)
Forested(5-25%)
Forested(>50%)
Peats
Podzolic Lithos olic
10
9
8
7
6
5
4
3
2
Control(<5%)
Forested(25-50%)
Fores ted(5-25%)
Forested(>50%)
Control(<5%)
Forested(25-50%)
Forested(5-25%)
Forested(>50%)
Poorly Drained Gleys
Median
25%-75%
Non-Outlier Range
Outliers
Extremes
Well Drained Mineral
Fig. 36: Box plot of the Margalef diversity Index scores at sites on various soil types on
sedimentary geology
Clustering of Biological, Chemical and Physical Metrics
The k-means algorithm was used to implement separate clustering of the 239 sites based on
selected chemical, physical and biological (macroinvertebrate response) metrics. In each
case the algorithm was asked to select four clusters.
50
The four biological clusters represented a gradient in the three metrics, group 1 having the
highest richness and cluster 4 the lowest. The latter contained a large proportion of the
impaired sites (31%, Table 12a). Thirty-three of these sites (44.5%) also grouped into the
chemical cluster (cluster 1, Table 13), the centroids of which represented the most acidic
conditions (Table 12b). Finally, 52 of the sites in the impoverished biological cluster 4
(66.6%, Table 13) also appeared in the physical cluster with the highest levels of percentage
coniferous cover (cluster 3) (Table 12c). The location of the biological cluster is shown in
Fig. 37. Sites within cluster 4 occurred in all regions.
Table 12: Centroid values for a) biological, b) chemical and c) physical clusters
(a)
Site
cluster1
cluster2
cluster3
cluster4
Taxon_Richness
42.53
38.61
33.23
25.59
Ephemeropteran_Abundance
108.13
465.11
55.91
37.75
Ephemeropteran_Richness
6.14
4.61
3.93
1.14
(b)
(c)
Site
cluster1
cluster2
cluster3
cluster4
%Coniferous
8.7
13.83
71.24
11.31
Slope
0.06
0.16
0.06
0.04
Elevation
297.36
292.02
187.81
113.4
51
Table 13: Numbers of sites loading into both a) biological and chemical clusters and b)
biological and physical clusters.
(a)
Biological
1
2
3
4
Biological
(b)
1
2
3
4
Chemical
1
2
1
19
33
Physical
1
11
4
12
8
2
10
9
7
5
2
6
3
5
5
3
11
1
14
16
3
17
6
20
52
4
29
6
20
20
4
20
5
21
13
Fig. 37: Location of the biological site clusters.
52
Multivariate Analyses
Much of the analyses presented thus far deals with trends in individual metrics.
The
multivariate analyses examined the relationships between the invertebrate communities at
each site.
To better visualise the similarities between sites in terms of community
composition NMDS plots, with Bray Curtis as the similarity measure, were generated using
the key indicator groups, Ephemeroptera, Plecoptera and Trichoptera, on site groups within
the selected geological settings. Plot were prepared for sites on peat and podzolic/lithosolic
soils draining granite and peat sites on sedimentary geology. These were highlighted earlier
as showing responses to forest cover.
Fig 38 is a plot of the sites draining peat on igneous/metamorphic geology. The control sites
are positioned largely to the left of the plot. The three most acidic control sites (DWW4,8 &
9 –Co. Wicklow; DG9-Co. Galway) with low taxon richness sit on the right side of the group
outline. At the other end of the plot sites M6, M7 and DWW6 (Knickeen, Co. Wicklow)
represent sites with high total taxon richness and good representation of Ephemeroptera
(richness and abundance) and Plecoptera. While there is, as expected, some overlap with the
most acidic controls, sites in the highest forest cover bands fall largely on the right side of the
plot. Those on the extreme right show impairment in a number of metrics.
The only
unimpaired sites within this region (DG21, 22 & 23-Owenwee River, Co. Galway) lie well
within the main grouping of control sites.
Fig. 38: NMDS plot of Bray Curtis similarity measure of EPT community composition at sites
draining peat on igneous/metamorphic geology. Outline of control sites excluded the acidic outliers.
53
A similar picture emerged with respect to sites draining podzolic/lithosolic on igneous
geology. Impaired sites within the three forestry bands plot on the right side of Fig. 39. In
contrast, the unimpaired afforested sites (DWW15 , 16 and 17) with high EPT taxon richness
and abundances plot to the left.
Fig. 39: MDS plot of Bray Curtis similarity measure of EPT community composition at sites draining
podzolic/lithosolic soils on igneous/metamorphic geology.
The control peat sites on sedimentary geology form a closer cluster than seen on igneous
geology (Fig. 40). Here sites with low EPT richness and abundance largely plot outside the
grouping of control sites, most of which show some impairment in the metrics applied earlier.
The frequency of impacted sites occurring outside of the control site grouping increases with
increasing forest cover.
Fig. 40: MDS plot of Bray Curtis similarity measure of EPT community composition at
sites draining peat on igneous/metamorphic geology.
.
54
Evaluation of the Degree of Biological Impairment
Five metrics (ephemeropteran richness, abundance Baetis spp., trichopteran richness,
evenness and diversity indices) which were not autocorrelated were selected to evaluate
potential impairment. The metrics selected target known indicator taxa as well as abundance,
evenness and diversity aspects of the community. As outlined in the methods section, sites
with metric values below two standard deviations of the mean of the control site values were
considered impaired. Table 14 shows the number of sites in each geological setting that
shows impairment for 1 to 5 metrics and gives an overall estimate of the number of sites that
fail on two or more metrics.
Table 14: Estimation of the number of impaired sites as indicated by low metric scores for 5
metrics in each geological and forest cover setting.
% Sites with
>2 impacted
sites
1
2
3
4
5
Total
Sites
22.2
33.3
5.6
22.2
16.7
22.2
0
16.7
16.7
33.3
16.7
11.1
0
0
27.8
9
7
17
55.6
50.0
77.8
0
11.1
25.0
75.0
11.1
25.0
0
33.3
37.5
0
22.2
0
0
0
0
4
9
8
*
66.7
62.5
0
27.3
6.7
25.0
18.2
10.0
50.0
9.1
20.0
0
27.3
33.3
0
9.1
20.0
4
11
30
*
63.6
83.3
Podzolic/Lithosolic
5-25%
25-50%
>50%
0
0
22.7
0
0
22.7
50.0
50.0
18.2
0
0
9.1
0
0
18.2
2
2
22
*
*
68.2
Poorly drained Gleys
5-25%
25-50%
>50%
0
14.3
9.1
33.3
28.6
18.2
0
0
0
0
0
18.2
0
0
18.2
3
7
11
*
28.6
54.5
Geological setting
Igneous/Metamorphic
Peat
5-25%
25-50%
>50%
Podzolic/Lithosolic
5-25%
25-50%
>50%
Sedimentary
Peat
5-25%
25-50%
>50%
•
low replication
55
The % impaired sites increased across the forest cover bands on igneous/metamorphic
geology. The same applied for peat sites on sedimentary geology. Podzolic/lithosolic sites
on sedimentary geology recorded significant impairment in the >50 forest band. It should
however be noted that replication was low in the other two forest bands. Some 55% of sites
on poorly drained gleys in the >50 forest band failed on more than two metrics. None of the
well drained mineral sites failed on more than two metrics.
Four sites recorded low
ephemeropteran abundances, however there was low replication of the control sites.
Comparison of Source and Downstream Communities
As outlined in the methods a number of forested streams were sampled above and below the
forest. Control sites were sampled at equivalent points. No significant differences were
detected between source and downstream sites in pairwise comparison statistics using
macroinvertebrate metrics (richness and abundance data for total taxa, Ephemeroptera,
Plecoptera, Trichoptera and Coleoptera). However, community differences were revealed by
the multivariate analyses.
Fig. 41 shows the NMDS plot based on Sorensen similarity measure. It clearly shows that
the control downstream sites (blue) were distinctly different from their sources (red). It also
indicates that the macroinvertebrate communities from forested downstream sites (green)
were similar to the sources from both forested (yellow) and control sites (red).
Fig. 41: MDS plot of Sorensen similarity measure of community composition at sites
upstream and downstream of forestry and at similar locations on control sites.
56
Longitudinal Variation in Macroinvetebrate Metrics
Several catchments were sampled at several sites to illustrate longitudinal changes in
macroinvertebrate community and to evaluate the potential distance downstream that a forest
effect might be detected. The example presented here is the King’s catchment, Co. Wicklow.
Some of the forested headwater sites had poor total taxon richness and ephemeropteran
richness compared to the control sites (Table 15) and this was maintained well down the
system to Site Kings1 (Fig. 42).
Fig. 42 Sites sampled along
the length of the King’s River,
Co. Wicklow.
Table 15: Locaton of sites sampled in the King’s catchment and recorded metric scores for
total taxon and ephemeropteran richness
River
% Forest
Taxon
Ephem.
Cover
Richness
Richness
ANNA3
0
25
2
Annalecka Brook
ANNA1
34.71
26
2
BALLIN2
0
24
5
Ballinagee River
BALLIN1
0
34
5
GLASH2
40.9
35
2
Glashaboy River
GLASH1
65.72
35
4
KINGS2
36.63
22
3
King’s River
KINGS1
28.26
25
2
Ephem.=Ephemeroptera
57
Further Evaluation of Potential Longitudinal Patterns in Macroinvertebrate
Recovery from Forest Effects
Forested sites in counties Wicklow (Vartry stream, DWW20 / VART1) and Cork (Foherish
river, DK26 / FOHER1) which recorded a paucity of Ephemeroptera during the spring 2007
sampling season were re-visited in spring 2008. In each case a nearby control/reference site
was also sampled, the Bohill river (DK23 / BOHIL1) in Co. Cork and the nearby nonforested tributary of the Vartry catchment, Co. Wicklow (DWW10 / VAR1). The controls
were selected to be comparable in terms of geology, soil type, elevation, catchment area,
slope, aspect and catchment size. The two paired streams were then sampled approximately
800 metres from source and every 500 metres thereafter over a two kilometre stretch of the
streams. The Cork sites drained areas of Old Red Sandstone, while all but one of the
Wicklow sites were situated on Palaeozoic sediments. Soils types within the catchment of
each paired stream were also comparable. The sampling sites where located between 200m360m asl. Sites 1 in all cases were located at >300m; Sites Nos. 2 between 275-300m; Sites
Nos. 3 between 250-275m; Sites Nos. 4 below 250m. The forested sites in Wicklow were
coded WKF1, WKF2, WKF3 and WKF4, while open (control) sites were labelled WKO1,
WKO2, WKO3 and WKO4. A similar site designation was used for the Cork sites (Figs. 43
& 44).
Fig. 43: Location of Wicklow forested
and control sites
Fig. 44: Location of Cork forested
and control sites.
58
The community composition was dominated by Chironomidae and other dipteran larva in the
Wicklow (WKO-64%; WKF-87%) and Cork sites
(CKO-57%; CKF-63%).
The
Ephemeroptera represeneted a lower percentage of the fauna at the Wicklow forested sites
(2%) compared to the control sites (10%). The Tricoptera varied little across sites ranging
from 4-7%.
The Plecoptera accounted for 15% (CKF) to 11% (CKO) of the total
abundances in the Cork sites compared to 2% (WKF) and 3% (WKO) sites. Crustaceans
were particularly abundant in non-forested sites; WKO and CKO accounting for 9% and 8%
of the total abundances. However, they accounted for only 4% of the fauna in forested sites
in Cork and where absent from forested sites in WKO. In Wicklow, non-forested sites
supported between 7 and 25 taxa, while forested sites had between 4 and 14. In Cork, nonforested sites recorded between 14 and 24 taxa, while between 8 and 21 taxa were found at
forested sites. The differences between the paired control and forested sites was significant
(Wicklow – Wilcoxon test: Z=-2.598, P<0.01; Cork – Wilcoxon test: Z=-2.096, P<0.05) (Fig.
45).
Taxon Richness
25
20
15
10
5
0
1
2
3
4
Site Number
Non-Forested Wicklow Sites
Forested Wicklow Sites
Taxon Richness
25
20
15
10
5
0
1
2
3
4
Site Number
Non-Forested Cork Sites
Forested Cork Sites
Fig. 45: Mean taxon richness at forested and non-forested sites in Wicklow and Cork. 59
Low ephemeropteran richness was a common feature of the first 2 sites in each forested
catchment (Figs. 46 & 47). The only species present was Baetis rhodani. While this
indicated a slight recovery in terms of species richness in comparison to the original sampling
period of the study (April-May 2007 - in which no mayfly were found), the abundances were
far lower than those for the corresponding controls in both Cork and Wicklow. Wilcoxon
paired Test, showed there was a significant differences in the mean ephemeropteran richness
between forested and non-forested sites in Cork (Z=-0.2366; P<0.05) and Wicklow (Z= 3.781; P<0.001). However, in Cork ephemeropteran richness differed significantly only
between the first three forested and control site pairs. The fourth, located almost 2.5km from
45).
.
the source, and still within close proximity to the forestry, was not showing impairment (Fig.
Ephemeropteran Richness
8
6
4
2
0
1
2
3
4
Site Number
Non-Forested Cork Sites
Forested Cork Sites
Fig. 46: Mean ephemeropteran richness at forested and non-forested sites in Cork
In Wicklow the significantly lower ephemeropteran richness at the forested sites persisted
down to Site 4. (Kruskall-Wallace ANOVA, P<0.05) (Fig. 47).
Trichopteran richness was
also significantly higher at the control sites in Wicklow (Z= -3.776’ P<0.001) and Cork (Z= 3.530, P <0.001) but here again the differences between the site pairs was eliminated by Site
4 in Cork.
60
Ephemeropteran Richness
8
7
6
5
4
3
2
1
0
1
2
3
4
Site Number
Non-Forested Wicklow Sites
Forested Wicklow Sites
Fig. 47: Mean ephemeropteran richness at forested and non-forested sites in Wicklow.
The longitudinal pH profile at the time of sampling is illustrated (Fig. 48). Although all sites
were circum-neutral there was at least a 0.5 unit of differences between the control and
forested site pairs and the differences was greatest for the first two sites.
8
7.5
Cork Non-Forested
7
Cork Forested
6.5
pH
Wicklow NonForested
6
Wickow Forested
5.5
5
1
2
3
4
Site No.
Fig. 48: Longitudinal pH profile for sites sampled in Wicklow and Cork.
61
3.3 FISH
As outlined in the methods sites were selected on a paired-catchment basis. To further ensure
that all site pairings were comparable, the coverage of each habitat type (riffle (P=0.38), glide
(P=0.402), pool (P=0.175)), depth (P=0.822), width (P=0.705), wetted area (P=0.812),
conductivity (P=0.492) and time fishing (effort – m2/min, P=0.12) were examined between
pairs. In each case there were no significant differences (Wilcoxon Ranked Sign Test,
P>0.05) detected, making all pairings comparable for further analysis.
The results of the Wilcoxon Signed Ranked Test on abundance of salmonids at the two site
grouping are given (Table 15). Overall the catch of trout was higher than salmon (Figs. 49 &
50), few salmon were caught in the Wicklow sites. The Galway sites recorded the highest
catches of salmonids. Total salmonid catch differed significantly between the paired control
and forested sites with the lowest numbers at the forested sites. Significant differences were
also detected for total salmon, trout and salmon fry (Table 16). In all cases, there were fewer
individuals at forested sites (Figs. 49 & 50). The differences were not significant for adult
trout or adult salmon.
Table 16: Wilcoxon Ranked Sign Test results for comparison of salmonid abundances
between non-forested and forested sites.
Total Salmonid Abundance
ZValue
-2.939
PValue
0.003*
Total Trout Abundance
Trout Adult Abundance
Trout Fry Abundance
-1.731
-1.168
-2.049
0.083
0.243
0.041*
Total Salmon Abundance
Salmon Adult Abundance
Salmon Fry Abundance
-2.194
-1.55
-2.366
0.028*
0.121
0.018*
* Significant to P = 0.05
Wilcoxon Sign Ranked Tests
62
Abundance
50
45
40
35
30
25
20
15
10
5
0
Total Salmon
Non-Forested Sites
Adult Salmon
Salmon Fry
Forested Sites
Abundance
Fig. 49: Mean abundances of salmon captured at forest and non-forested sites. Standard error
bars included.
50
45
40
35
30
25
20
15
10
5
0
Total Trout
Non-Forested Sites
Adult Trout
Trout Fry
Forested Sites
Fig. 50: Mean abundances of trout captured at forest and non-forested sites. Standard error
bars included.
Salmonid densities (fish/m2) were also compared between the paired control and forested
sites. The results were similar to those described for fish catch. Total trout density (Z = 2.45,
P = 0.014) and trout fry (Z = 2.50, P = 0.0122) density differed between the two site groups.
63
The differences were significant for adult trout. Total salmon (Z = 2.73, P = 0.006) and fry
(Z = 3.54, P = 0.0003) density was significantly lower in the forested sites.
The length frequency distribution of salmonids across all sites is shown in Fig. 51. The
populations were generally dominated by 1+ fish ranging in length from 9 to 15 cm. The
numbers of larger fish were highly variable across sites. Fry numbers were generally lower at
the forested sites. An evaluation of growth differences between control and forested sites
was undertaken for fry. The data were examined on a regional basis to incorporate any
natural variation in growth. No significant differences were detected in any of the regions or
across the total dataset (Mann-Whitney, P>0.05, Table 17).
Table 17: Mann-Whitney test statistic results (P>0.05) for a comparison of the length of trout
fry in control and forested sites in four regions.
Region
Wicklow
Mayo
Galway
Donegal
Mann-Whitney Test
Statistic
0.260
0.279
0.314
0.082
Several sites were noted for their low abundances or paucity of salmonids (Table 18). These
sites were all highly forested with the exception of one control site on the Cloghoge river in
Co. Wicklow (CLOG1). Kelly-Quinn et al. (1996b) suggest that the combination of site
elevations >400 m.a.s.l along with steep slopes can exclude salmonid fish from streams.
However, while several of the sites in Table 18 were approaching the elevation cut-off, the
slopes were not steep. None of the sites had any known barriers to fish movement.
64
a)
c)
b)
d)
Fig. 51: Fish population structure of the salmonids caught in the control (blue) and forested
(green) sites in the Galway (a), Mayo (b), Donegal (c) and Wicklow (d) regions. The stream
pairs are maintained alongside each other for comparative purposes.
65
Table 18: Sites with absences of salmonid fish.
All Fish Absent
ANNA1 (25-50% Forest)
Fry Absent
ANNA1 (25-50% Forest)
GAMON5 (>50% Forest)
INCH1 (25-50% Forest)
GMOY1 (>50% Forest)
GLASH2 (25-50% Forest)
CORRIB1 (>50% Forest)
CORRIB2 (25-50% Forest)
SRAG1 (5-25% Forest)
CLOG1 (Control <5%
Forest)
Adult Fish Absent
ANNA1 (25-50% Forest)
CROE1 (5-25% Forest)
66
4. DISCUSSION
This project set-out to investigate the presence and extent of any acidification associated with
coniferous forestry in Ireland and to assess the risk of impact with respect to different
geological settings. In the selection of forested sites it was aimed to represent a combination
of the risk factors in terms of catchment cover and acid-sensitive geology that were perceived
to have the greatest potential for acidification. The large number of sites selected allowed for
good spatial coverage but it did limit the amount of water sampling that could be undertaken.
The aim was to sample each site at variable flow conditions, from low flow to flood. It was
however not possible to obtain flood samples for all sites as a result of their geographic
spread and remote locations. Furthermore, it was often difficult to ascertain the stage in the
hydrograph represented on any one date.
Nevertheless, within any one region a good
representation of control and forested sites were sampled within the same timeframe and
usually under the same flow/weather conditions.
The pH results analyses suggested that most of the streams were episodically acidic with a
small group more likely to be circum-neutral. Overall, the pH results indicated increased
acidity at some sites associated with forestry on peat and podzolic/lithosoilic soils on both
igneous/metamorphic and sedimentary geology and to some extent on poorly drained gleys.
Two components of these results require clarification. Firstly, while a small number of the
control sites, especially on peat/granite, recorded minimum pH values as low as some of the
forested sites the frequency of low pH readings was substantially higher among some groups
of forested sites. So the critical issue may be that the frequency and duration of acid pulses
can be higher in some forested catchments. Previous intensive monitoring of acid pulses in a
heavily afforested stream in the Wicklow mountains alluded to this (Kelly-Quinn, Tierney &
Bracken, 1997). The current dataset, unfortunately, does not have sufficient data to further
test this hypothesis.
This association emerged when forest cover in the site catchment
exceeded 25-30%. Factors controlling the severity and duration of acid pulses require further
research to better target measures.
The second issue to consider is whether the current
results suggest a forest-cover threshold above which the risk of acidification increases.
Certainly
the
minimum
pH
for
both
peat
and
podzolic/lithosolic
sites
on
igneous/metamorphic geology began to fall below the lower limit of the control sites when
forest cover exceeded values in the region of 25%. The same applied to peat sites draining
sedimentary geology. Sites on podzolic/lithosolic soils on sedimentary geology did not have
67
minimum pH values below the lower limit of the control sites until forest cover exceeded
60%. A similar threshold might be applied to sites on poorly drained gleys but the level of
replication is too low for this decision.
It is important to point out that not all sites within the high forest-cover bands had low
minimum pH.
It may be that the sampling did not capture the peak of the acidity or
alternatively the sites are buffered against pH change. Indeed, when alkalinity was examined,
many sites, particularly those on peat/podzolic/lithosolic on sedimentary geology, had
alkalinity values well above 20 mg/l CaCO3. Further analyses of 57 of these sites confirmed
that 78% have some alkaline sub-soils or carbonate geology within the catchment.
Interestingly, a number of sites with maximum alkalinity >20 mg/l CaCO3 recoded high flow
alkalinity values of close to zero. This was also mirrored in the SDI results for these sites.
Evaluation of the differences in the flow pathway between base and flood conditions in
forested catchments is clearly required to better understand factors controlling buffering
potential. Overall, the greatest variation in alkalinity was recorded on sedimentary geology
which may relate to more complex geology with, as already mentioned, occurrences of some
carbonate soils or rocks among the largely acid-sensitive geology. More detailed spatial and
temporal analyses of the chemical characteristics of waters draining sedimentary geology is
required for more precise mapping of acid sensitivity and this should be an element of future
research projects.
The presence of forestry tended to depress site pH and alkalinity. Calculations suggested that
dilution makes a variable contribution to loss of alkalinity and in many cases the forested
sites showed a slightly higher % value. Anion titration was detected in all events examined.
The principal contributors were organic acids and sulphate. Excess sulphate only made a
contribution in the Wicklow sites and at one site in Galway. The contribution of nitrate
across all sites was insignificant. The contribution of sea salts to acidification was low and
only one significant sea-salt event was detected at one site in Galway. Similar variability in
contributing variables has been reported by Kowalik et al. (2007). Overall, it is likely that a
combination of dilution and higher organic acidity concentrations, and occasionally excess
sulphate, contributed mostly to the differences in acidity between control and heavily
afforested catchments. Reasons for the differences in organic acidity are unclear and may
relate to the effects of patterns of drying and wetting and other climatic factors associated
with forest soils (Raveh and Avnimelech, 1978; Worrall, Burt and Adamson, 2004). This
68
represents another key knowledge gap and if addressed may help to develop focused forestry
practices that minimize the risk of acidification. Indeed, given that the focus of the current
study was on mature forestry, we need to determine the acidification risk associated with
each of the key forestry practices from site preparation to felling.
In terms of the macroinvertebrtae analyses the control sites draining sedimentary sites were
more productive in terms of biomass than the igneous/metamorphic sites but the ranges for
taxon richness were similar.
The higher total macroinvertebrate abundances at the
sedimentary sites could be largely attributed to the Ephemeroptera and Chironomidae. This
may relate to slightly higher pH and cation concentrations in the former. In both geological
settings the Ephemeroptera were reduced in abundance at sites in the two highest forest cover
bands, a factor of increasing pH. The sensitivity of the Ephemeroptera to acidification is well
established through field observations as well as stream microcosm experiments (e.g.
Courtney &Clements, 1998) Interestingly, at the igneous/metamorphic sites the reduction in
ephemeropteran abundance was largely balanced by an increase in the numbers of Plecoptera.
This did not occur at the sites draining sedimentary geology and consequently overall
abundance declined gradually across the forest cover bands. Most the analyses in relation to
forestry effects was performed separately for the two geological settings to avoid any
confounding effects of differences in taxon abundances.
Overall, the biological data largely mirrored the trends for the acidity variables. In fact pH
was the key variable structuring the community, a feature that is commonly reported in the
literature (Ormerod and Edwards, 2006; Sutcliffe and Carrick, 1973). The study on the
source communities illustrated two key findings. Firstly, the macroinvertebrate communities
found in control source streams differ substantially from those at sites located further
downstream and therefore one must be cautious about making comparisons between sites
upstream and downstream of forestry if source sites are included. Secondly, the study
showed that the macroinvertebrates communities downstream of the forest were more similar
to the source sites than the downstream counterpart on the non-forested stream. This implies
some impact of forestry inputs, most likely relating to acidity.
Several metrics (taxon richness, ephemeropteran richness, abundance of baetids, EPT
richness, diversity indices), which showed a strong relationship with pH, were also shown to
vary significantly across the forest cover bands or to correlate with % forest cover bands.
69
Ephemeroptera were absent from several sites in the >50 forest bands on peat and
podzolic/lithosolic soils.
A striking finding was that the number of sites with low
ephemeropteran richness and abundance increased across the forest cover bands.
The
analyses on the individual metrics highlighted similar % forest thresholds for risk of impact
as described for the hydrochemistry.
When a selection of non-correlated metrics were
combined it was clear that a large proportion of sites in the >50% cover band, and a smaller
number of the 25-50% band, had some degree of impairment. These same sites were shown
to have a different invertebrate community to the control sites as indicated on the NMDS
plots.
The implications of the detected impairment for overall ecological health and
functioning requires further research.
It should be highlighted that, as for pH, not all sites within the high forest cover bands
showed impact. Most of those that were impacted recorded minimum pH values well below
5.0 and alkalinity values below or close to zero but this was not consistent across the
soil/geological categories. In fact, some of the non-impacted sites also became substantially
acidic.
Further research must target these sites to better understand the mechanisms
governing responses to acid impact under naturally acidic conditions. It is likely that both the
degree and duration of acidity are important factors.
Some limited analyses of the season and longitudinal extent of forest impact was undertaken.
The results suggest that impact may be seasonal and that recovery in some of the metrics (e.g.
baetid numbers) takes places. The potential for seasonal recovery may be dependent on
climatic factors viz. severity, duration and frequency of precipitation leading to acid pulses;
but also life history patterns of the biota (e.g. baetids). The limited data also suggest that
recovery may occur over a shorter distance on sedimentary geology than on
igneous/metamorphic geology but this is likely to be controlled by many interacting factors
that change with distance from the source, such as catchment size, forest cover, inputs from
other sub-catchments as well as geology/soils. The longitudinal responses to forestry as a
land-use activity needs to be addressed by further research.
The fish analyses was limited to 19 paired sites with similar habitat but did highlight
significant differences in fish catch and density between the control and forested groups.
This difference was mainly attributed to low numbers of fry (salmon and trout) in the forested
streams. Low recruitment in forested streams is most likely to be related to pH as previously
70
highlighted by Kelly-Quinn, Tierney and Bracken (1993) in upland streams in Co. Wicklow.
However, there may be other contributing factors such as discharge and availability of food.
In conclusion this study has addressed the objectives as set out in the introduction. It has
detected increases in acidity and biological impairment associated with forest plantations.
The risk appears to be a factor of soil geology/soil and increasing catchment cover with the
greatest impact occurring above a coniferous forest cover of 50%. The sources and pathways
of acid inputs needs to be better clarified and related to forest activities to allow further
refinement of the programme of measures.
5. REFERENCES
Allott, N., Mills, W.R.P., Dick, J.R.W., Eacrett, A.M., Brennan, M.T., Clandillon, S.,
Phillips, W.E.A., Critchley, M., and Mullins, T.E. (1990) Acidification of Surface
Waters in Connemara and South Mayo. Current Status and Causes, Dublin:
duQuesne Limited, Economic and Environmental Consultants, pp 61.
Allott, N., Brennan, M., Cooke, D., Reynolds, J., and Simon, N. (1997) Stream chemistry,
hydrology and biota, Galway-Mayo region. In: A study of the effects of stream
hydrology
and
water
chemistry
in
forested
catchments
on
fish
and
macroinvertebrates, AQUAFOR Report 4, COFORD, Dublin, Ireland.
Anon. (2005) The characterisation and analysis of Ireland’s river basin districts in
accordance with Section 7(2 & 3) of the European Communities (Water Policy)
Regulations 2003 (SI 722 of 2003). National Summary Report (Ireland).
Arnason, F., Antonsson, T. and Einarsson, S.M. (2005) Evaluation of single-pass electric
fishing to detect changes in population size of Atlantic salmon (Salmo salar L.)
juveniles. Icelandic Agricultural Sciences, 18: 67-73.
Barbour, M.T., Gerritsen, J. and Snyder, B.D. (1997) Revision to Rapid Bioassessment
Protocols for Use in Streams and Rivers. EPA 841-D-97-002, Office of Water, United
States Environmental Protection Agency, Washington, D.C., USA.
Bertrand, K.N. Gido, K.B. and Guy, C.S. (2006) An evaluation of single-pass versus
multiple-pass backpack electrofishing to estimate trends in species abundance and
richness in prairie streams. Transactions of the Kansas Academy of Science, 109: 131138.
71
Bowman, J. (1991) Acid sensitive surface waters in Ireland. Environmental Research Unit,
Dublin. pp 321.
Clenaghan, C., Giller, P.S., O'Halloran, J., and Hernan, R. (1998) Stream macroinvertebrate
communities in a conifer-afforested catchment in Ireland: relationships to physicochemical and biotic factors. Freshwater Biology 40: 175-193.
Courtney, L.A. and Clements, W.H. (1998) Effects of acidic pH on benthic macroinvertebrate
communities in stream microcosms. Hydrobiologia 379: 135-145.
Economou, A.N., Schmutz, S., Melchior, A. and D. Heidvogel (2002) Development,
evaluation & implementation of a standardised fish-based assessment method for the
ecological status of European rivers: A contribution to the Water Framework
Directive (FAME): Defining reference conditions (D3), Final Report. FAME, Austria.
Evans, C.D., Reynolds, B., Hinton, C., Hughes, S., Norris, D., Grant, S. and Williams, B.
(2008) Effects of decreasing acid deposition and climate change on acid extremes in
an upland stream. Hydrology and Earth System Sciences 12: 337-351.
Farrell, E.P. (1995) Atmospheric deposition in maritime environments and its impact on
terrestrial ecosystems. Water, Air, & Soil Pollution, 85: 123-130.
Giller, P., O'Halloran, J., Kiely, G., Evans, J., Clenaghan, C., Hernan, R., Roche, N., and
Morris, P. (1997) An Evaluation of the Effects of Forestry on Surface Water Quality
and Ecology in Munster. In A study of the effects of stream hydrology and water
quality in forested catchments on fish and invertebrates AQUAFOR Report 2,
COFORD, Dublin, Ireland: COFORD, pp 77.
Harriman, R., and Morrison, B.R.S. (1982) Ecology of streams draining forested and nonforested catchments in an area of central Scotland subject to acid precipitation.
Hydrobiologia, 88: 251-263.
Harriman, R., Anderson, H., and Miller, J.D. (1995) The role of sea-salts in enhancing and
mitigating surface water acidity. Water, Air, & Soil Pollution, 85: 553-558.
Hartigan, J.A., Wong, M.A., (1979) Algorithm AS 139: A k-means clustering algorithm.
Applied Statistics 28: 100-108.
Hayes. J.W. and Baird, D.B. (1994) Estimating relative abundance of juvenile brown trout in
rivers by underwater census and electrofishing. New Zealand Journal of Marine and
Freshwater Research, 28: 243-253.
Hornung, M., Le-Grice, S., Brown, N., and Hughes, S. (1990) The role of geology and soils
in controlling surface water acidity in Wales. In Acid Waters in Wales (R.W.
72
Edwards, J.H. Stoner, and A.S. Gee, eds.), Dordrecht: Kluwer Academic Publishers,
pp 55-66.
Hornung, M., Bull, K.R., Cresser, M., Ullyett, J., Hall, J.R., Langan, S., Loveland, P.J., and
Wilson, M.J. (1995) The sensitivity of surface waters of Great Britain to acidification
predicted from catchment characteristics. Environmental Pollution, 87: 207-214.
Jenkins, A., Cosby, B.J., Ferrier, R., Walker, T.A.B., and Miller, J.D. (1990) Modelling
stream acidification in afforested catchments. An assessment of the relative effects of
acid deposition and afforestation. Journal of Hydrology, 120: 163-181.
Jones, M.L. and Stockwell, J.D. (1995) A rapid assessment procedure for the enumeration of
salmonine populations in streams. North America Journal of Fisheries Management,
15: 551-562.
Johnson, M.J., Giller, P.S., O’Halloran, J., O’Gorman, K. and Gallagher, M.B. (2005) A
novel approach to assess the impact of landuse activity on chemical and biological
parameters in river catchments. Freshwater Biology, 50: 1273–1289.
Kahl, J.S., Norton, S.A., Haines, T.A., Rochette, E.A., Heath, R.H. and Nodvin, S.C. (1992)
Mechanisms of episodic acidification in low-order streams in Maine, USA.
Environmental Pollution 78: 37-44.
Kelly-Quinn, Tierney, D. and Bracken, J.J. (1993) Survival of salmon eggs planted in upland
streams. Aquaculture and Fisheries Management 24: 791-796.
Kelly-Quinn, M., Tierney, D., Coyle, C. and Bracken, J.J. (1996a) Factors affecting the
susceptibility of Irish soft-water streams to forest-mediated acidification. Fisheries
Management and Ecology 3, 287-301.
Kelly-Quinn, M., Tierney, D., Roche, W. and Bracken, J.J. (1996b) Distribution and
abundance of trout populations in moorland and afforested upland nursery streams in
Co. Wicklow. Biology and Environment, Proceeding of the Royal Irish Academy. 96B,
No. 3 127-140.
Kelly-Quinn, M., Tierney, D. and Bracken J.J. (1997) Stream Chemistry, Hydrology and
Biota, Wicklow Region. In: A study of the effects of stream hydrology and water quality
in forested catchments on fish and invertebrates. AQUAFOR Report 3, COFORD,
Dublin. 92pp. ISBN:0 9523938 4 0
Kowalik, R.A., Cooper, D.M., Evans, C.D. and Ormerod, S.J. (2007) Acidic episodes retard
the biological recovery of upland British streams from chronic acidification. Global
Change Biology 13: 2439-2452.
73
Kruse, C.G., Hubert, W.A. and Rahel, F.J. (1998) Single-pass electrofishing predicts trout
abundance in mountain streams with sparse habitat. North American Journal of
Fisheries Management, 18: 940-946.
Mitro, M.G. and Zale, A.V. (2000) Predicting fish abundance using single-pass removal
sampling. Canadian Journal of Fisheries and Aquatic Sciences, 57: 951-961.
Ormerod, S.J., Wade, K.R., and Gee, A.S. (1987) Macrofloral assemblages in upland Welsh
streams in relation to acidity, and their importance to invertebrates. Freshwater
Biology, 18: 545–557.
Ormerod, S.J., and Wade, K.R. (1990) The role of acidity in the ecology of Welsh lakes and
streams. In Acid Waters in Wales (R.W. Edwards, J.H. Stoner, and A.S. Gee, eds.),
Dordrecht: Kluwer Academic Publishers, pp 95-109.
Ormerod, S.J., Rutt, G., Weatherley, N.S., and Wade, K. (1991) Detecting and managing the
influence of forestry on river systems in Wales: Results from surveys, experiments
and models. In: Irish Rivers: Biology and Management (M.W. Steer, ed.) Royal Irish
Academy, Dublin. pp 163-184.
Ormerod, S.J. and Edwards, R.W. (2006) The ordination and classification of
macroinvertebrate assemblages in the catchment of the River Wye in relation to
environmental factors. Freshwater Biology, 17: 533-546.
Raveh, A. and Avnimelech, Y. (1978) The effect of drying on the colloidal properties and
stability of humic compounds. Plant Soil, 50: 545-552.
Rees, R.M., and Ribbens, J.C.H. (1995) Relationships between afforestation, water chemistry
and fish stocks in an upland catchment in south west Scotland. Water, Air, & Soil
Pollution, 85: 303-308.
Resh, V.H., Brown, A.V., Covich, A.P., Gurtz, M.E., Li, H.W., Minshall, G.W., Reice, S.R.,
Sheldon, A.L., Wallace, J.B. and Wissmar, R.C. (1988) The role of disturbance in
stream ecology. Journal of the North American Benthological, 7: 433-455.
Stoner, J., Gee, A., and Wade, K. (1984) The effects of acidification on the ecology of
streams in the upper Twyi catchment in West Wales. Environmental Pollution, A, 36:
125-157.
Sutcliffe, D.W. and Carrick, T.R. (1973) Studies on mountain streams in the English Lake
District I. pH, calcium and the distribution of invertebrates in the River Duddon,
Freshwater Biology, 3:, 437-462.
74
Tierney, D., Kelly-Quinn, M., and Bracken, J.J. (1998) The faunal communities of upland
streams in the eastern region of Ireland with reference to afforestation impacts.
Hydrobiologia, 389: 115-130.
Waters, D., and Jenkins, A. (1992) Impacts of afforestation on water quality trends in two
catchments in mid-Wales. Environmental Pollution, 77: 167-172.
Worrall, F., Burt, T. and Adamson, J. (2004) Can climate change explain increases in DOC
flux from upland peat catchments? Science of the Total Environment, 326: 95-112.
Wright, J.F. (1995) Development and use of a system for predicting the macroinvertebrate
fauna in flowing waters. Australian Journal of Ecology, 20: 181-197.
75
APPENDIX A
Location of the 239 sampling sites
76
Main system
Trib of River Blackwater
Caher River
Dripsey River
Dripsey River
Trib of Caher River
Glennaharee River
Glengarriff Stream
Fermoyle River
Trib of River Blackwater
Ownagluggin River
Ownagluggin River
Ownagluggin River
Ownagluggin River
Carrigduff River
Carrigduff River
Crinnaloo River
Aghalode River
Trib of River Laney
Trib of River Blackwater
Bregoge River
Castlepook River
Trib of Bregoge River
Fluckane Stream
Trib of River Funshion
Trib of Sheep River
Garrane River
Trib of Ross River
Trib of River Bride
Bunnaglanna River
Ballycorban River
Trib of Scarriff River
Bow River
Bow River
Corlea River
Trib into Lough Atorick
Muchnagh
Douglas
Muchnagh
Araglin
Araglin
Araglin
Geeragh River
Geeragh River
Burncourt River
Trib Araglin
Glenakeefe
Glennandaree
Sheep
Invertebrate
Site Code
CK1
CK10
CK12
CK13
CK14
CK15
CK17
CK19
CK2
CK21
CK22
CK23
CK24
CK25
CK26
CK27
CK28
CK29
CK3
CK30
CK31
CK32
CK33
CK34
CK35
CK36
CK4
CK6
CK8
CL2
CL3
CL4
CL5
CL6
CL8
MUCH1
DOUG3
MUCH2
ARAG1
ARAG2
ARAG3
GEER1
GEER2
BURN1
ARA1
GKEEF1
GLENN1
SHEEP1
Water Chemistry Site
Code
CK1
CK10
CK12
CK13
CK14
CK15
CK17
CK19
CK2
CK21
CK22
CK23
CK24
CK25
CK26
CK27
CK28
CK29
CK3
CK30
CK31
CK32
CK33
CK34
CK35
CK36
CK4
CK6
CK8
CL2
CL3
CL4
CL5
CL6
CL8
DC1
DC10
DC12
DC13
DC14
DC15
DC16
DC17
DC18
DC2
DC4
DC5
DC6
County
Cork
Cork
Cork
Cork
Cork
Cork
Cork
Cork
Cork
Cork
Cork
Cork
Cork
Cork
Cork
Cork
Cork
Cork
Cork
Cork
Cork
Cork
Cork
Cork
Cork
Cork
Cork
Cork
Cork
Clare
Clare
Clare
Clare
Clare
Clare
Cork/Tipperary
Cork/Tipperary
Cork/Tipperary
Cork/Tipperary
Cork/Tipperary
Cork/Tipperary
Cork/Tipperary
Cork/Tipperary
Cork/Tipperary
Cork/Tipperary
Cork/Tipperary
Cork/Tipperary
Cork/Tipperary
Grid
Reference
W 717 958
W 452 863
W 416 857
W 415 858
W 444 869
W 459 889
W 454 922
W 394 919
W 671 973
W 370 872
W 377 875
W 377 875
W 384 877
W 357 884
W 354 883
W 370 892
W 384 851
W 353 855
W 666 970
R 595 134
R 606 137
R 620 133
R 631 131
R 693 140
R 720 149
R 594 179
W 647 968
W 740 926
W 709 928
R 635 891
R 639 891
R 663 904
R 669 917
R 616 938
R 641 940
R 882 071
R838 057
R 868 069
S 007 066
S 006 067
S 006 068
R 824 178
R 829 193
R 937 192
R 904 062
S 062 060
S 040 077
R 910 178
Easting
171785
145243
141683
141503
144463
145995
145465
139459
167196
137010
137788
137749
138438
135710
135486
137044
138427
135399
166688
159550
160600
162013
163110
169333
172068
159451
164725
174075
170926
163510
163978
166353
166934
161652
164190
188278
183827
186850
200701
200602
200662
182482
182988
193779
190438
206269
204056
191095
Northing
95834
86306
85758
85862
86909
88996
92232
91961
97330
87248
87521
87501
87701
88412
88331
89243
85148
85576
97055
113454
113785
113334
113186
114043
114983
117961
96815
92612
92860
189122
189115
190427
191795
193837
194005
107108
105741
106913
106648
106743
106800
117850
119382
119290
106297
106014
107794
117840
77
Main system
Sheep
Sheep
Douglas
Gweebarra River
Gweebarra River
Elatagh River
Elatagh River
Elatagh River
Elatagh River
Trib Deele River
Stranagoppoge
Gweebarra River
Gweebarra River
Sruhanboy River
Trib of Strachashell River
Cloghroe River
Stranagoppoge
Trib Owenmore
Trib. Owenree
Owenboliska
Owenboliska
Owenboliska
Owenboliska
Owenboliska
Lough More
Owenboliska
Sruffaunanulra River
Trib Bunowen River
Trib Lough Corrib
Owenwee River
Owenwee River
Owenwee River
Trib to Maumwee Lough
Trib to Owenriff
Glengawbeg River
Glengawbeg River
Trib Owenmore
Owenakilla River
Gowlaun River
Trib Owenwee
Trib Owenwee
Loughanillaunmore
Owenboliska
Owendunnakilla
Knockbane river
River Loo
Clydagh
Clydagh
Clydagh
Clydagh
Invertebrate
Site Code
SHEEP3
SHEEP2
DOUG2
GBAR1
GBAR3
ELATA2
ELATA4
ELATA1
ELAT3
DEEL1
STRAN2
GBAR2
GBAR4
SRUHA1
STRAC1
CROE1
STRAN1
OWEN2
OREE1
OLISKA2
Oliska1
OLISKA7
OLISKA6
OLISKA5
MORE1
OLISKA4
SRUFF1
BOWEN1
CORRIB1
OWEE1
OWEE2
OWEE3
MAUM1
ORIFF2
GBEG2
GBEG3
OWEN3
OWENK1
GLAUN1
OWENN2
OWENN1
LOUGH1
OLISKA3
OKILLA1
KBANE1
LOO1
CLYDA9
CLYDA10
CLYDA6
CLYDA1
Water Chemistry Site
Code
DC7
DC8
DC9
DD10
DD11
DD13
DD14
DD15
DD16
DD2
DD3
DD4
DD5
DD6
DD7
DD8
DD9
DG1
DG11
DG12
DG13
DG14
DG15/G7
DG16
DG17
DG18
DG19
DG2
DG20
DG21
DG22
DG23
DG24
DG25
DG28
DG29
DG3
DG30
DG31
DG4
DG5
DG6
DG7/G6
DG8
DG9
DK1
DK11
DK12
DK13
DK14
County
Cork/Tipperary
Cork/Tipperary
Cork/Tipperary
Donegal
Donegal
Donegal
Donegal
Donegal
Donegal
Donegal
Donegal
Donegal
Donegal
Donegal
Donegal
Donegal
Donegal
Galway
Galway
Galway
Galway
Galway
Galway
Galway
Galway
Galway
Galway
Galway
Galway
Galway
Galway
Galway
Galway
Galway
Galway
Galway
Galway
Galway
Galway
Galway
Galway
Galway
Galway
Galway
Galway
Cork/Kerry
Cork/Kerry
Cork/Kerry
Cork/Kerry
Cork/Kerry
Grid
Reference
R 906 201
R 893 203
R 850 056
B 857 044
B 861 026
C 043 039
C 045 041
C 043 039
C 022 052
C 112 031
G 912 976
B 854 038
B 840 005
C 048 017
G887 965
C102 009
G 924 992
L 913 731
M 015 468
M 142 360
M 145 355
M 084 322
M 085 327
M 080 332
M 068 310
M 104 345
M 090 379
L 837 757
M 057 485
M 025 452
M 032 455
M 033 458
L 973 484
M 083 423
M 056 409
M 053 410
L 929 728
M 097 465
M 089 471
L 951 771
L 945 771
M 098 283
M 112 345
M 165 364
M 171 353
W 048 796
W 176 843
W165 842
W 159 829
W 183 833
Easting
190631
189384
185087
185721
186163
204371
204500
204307
202224
211241
191262
185497
184082
204803
188777
210272
192448
91317
101571
114238
114582
108464
108519
108003
106835
110492
109007
83779
105770
102550
103259
103301
97370
108363
105672
105371
92970
109748
108947
95168
94573
109823
111290
116552
117126
104813
117667
116554
115954
118389
Northing
120168
120327
105620
404452
402664
403953
404109
403947
405222
403177
397661
403876
400536
401716
396441
400935
399263
273196
246870
236010
235506
232252
232725
233271
231025
234516
237978
275767
248510
245210
245508
245877
248412
242381
240995
241083
272821
246534
247150
277166
277143
228341
234524
236440
235361
79671
84397
84248
82944
83368
78
Main system
Clydagh
Roughty River
Roughty River
Roughty River
Clydagh
River Loo
Clydagh
Owgarriv River
Bohill River
Clydagh
Foherish River
Inchamore River
Aughboy River
Slievenaneav River
Roughty River
Trib to Flesk River
Roughty River
Inchamore River
Kealgorm
Kealgorm
Trib to Slaheny River
Garrrow River
Glenthomas River
Glennamong River
Trib Srahmore River
Trib Skerdagh River
Trib Crumpaun River
Trib Crumpaun River
Fiddaungrave
Fiddaungal
Glennamong River
Glenthomas River
Glendahurk River
Glendahurk River
Glendahurk River
Fiddaunatoreen
Glennamong River
Glennamong River
Vartry
Derrybaun River
Annalecka Brook
Glashaboy
Garryknock
Oiltiagh Brook
Inchavore River
Ballinagee River
Vartry
Lugduff Brook
Lugduff Brook
Invertebrate
Site Code
CLYDA2
ROUGH3
ROUGH1
ROUGH2
CLYDA3
LOO2
CLYDA8
OWGAR1
BOHIL1
CLYDA7
FOHER1
INCHMR1
AUGHB1
SLIEVE1
ROUGH5
FLESK1
ROUGH4
INCHMR2
KEAL1
KGORM1
SLAH1
GARW1
GTHOM1
GAMON2
SRAH1
SKERD1
CRUM2
CRUM1
FIDD2
FIDD1
GAMON3
GTHOM2
GHURK3
GHURK1
GHURK2
FREEN1
GAMON1
GAMON4
VAR1
DERRY1
ANNA1
GLASH1
GARRY1
OILI1
INCH1
BALLIN2
VART1
LUG3
LUG1
Water Chemistry Site
Code
DK15
DK16
DK17
DK18
DK19
DK2
DK20
DK22
DK23
DK24
DK26
DK27
DK28
DK29
DK3
DK30
DK31
DK32
DK4
DK5
DK6
DK8
DM1
DM10
DM11
DM12
DM13
DM14
DM15
DM16
DM17
DM2
DM3
DM4
DM5
DM6
DM7
DM8
DWW10
DWW12
DWW13
DWW15
DWW16
DWW17
DWW19
DWW2
DWW20
DWW21
DWW22
County
Cork/Kerry
Cork/Kerry
Cork/Kerry
Cork/Kerry
Cork/Kerry
Cork/Kerry
Cork/Kerry
Cork/Kerry
Cork/Kerry
Cork/Kerry
Cork/Kerry
Cork/Kerry
Cork/Kerry
Cork/Kerry
Cork/Kerry
Cork/Kerry
Cork/Kerry
Cork/Kerry
Cork/Kerry
Cork/Kerry
Cork/Kerry
Cork/Kerry
Mayo
Mayo
Mayo
Mayo
Mayo
Mayo
Mayo
Mayo
Mayo
Mayo
Mayo
Mayo
Mayo
Mayo
Mayo
Mayo
Wicklow
Wicklow
Wicklow
Wicklow
Wicklow
Wicklow
Wicklow
Wicklow
Wicklow
Wicklow
Wicklow
Grid
Reference
W 222 845
W 068 751
W 072 712
W 065 710
W 210 865
W 045 782
W 201 844
W 100 800
W 195 806
W 206 845
W 241 807
W117 775
W 125 783
W 089 800
W 038 743
W 102 842
W 097 730
W 120 773
W 012 771
W 007 782
W 029 700
W 081 753
L 889 999
F 928 038
F 965 054
G 012 023
G 046 021
G 073 047
G 061 069
G 055 076
F 941 037
F 887 003
F 910 009
L 909 985
F 912 007
F 950 019
F 944 027
F938 304
O 204 092
T 133 946
O 067 026
O 053 013
O 026 022
S 991 958
O 110 060
O 047 047
O 190 069
T 110 955
T 111 957
Easting
122212
106817
107266
106572
121066
104554
120194
110094
119594
120620
124186
111751
115953
108969
103836
110254
109706
112095
101263
100724
102958
108180
88932
92869
96560
101267
104638
107325
106136
105540
94103
88783
91083
90982
91275
95086
94493
93899
320426
313316
306721
306599
302636
299141
311100
304702
319044
311096
311123
Northing
84523
75194
71289
71000
86565
78231
84468
80028
80696
84561
80767
77576
78398
80033
74388
84282
73061
77334
77185
78269
70085
75351
299925
303830
305240
302346
302162
304771
306948
307619
303724
300376
300953
298562
300780
301904
302799
304237
209227
194649
202647
201610
202250
195835
206200
204744
206912
195572
195749
79
Main system
Sraghoe Brook
Glashaboy
Annalecka Brook
Knickeen River
Slaney River
Cloghoge River
Cloghoge River
Trib of Owendalulleegh
River
*
Trib of Boleyneendorrish
River
Trib of Boleyneendorrish
River
Trib into Derryclare Lough
Trib into Derryclare Lough
Trib of Owenglin River
Trib of Owendalulleegh
River
Owenaglanna River
*
Owendunnakilla
Trib of Owenboliska River
*
Trib of Owenboliska River
Trib into Seecon Lough
Trib of Smearlagh River
Dromaddamore River
Trib of Smearlagh River
Barranahown River
Trib of Ahaphuca River
Trib of Awbeg River
Trib of Awbeg River
Trib of Awbeg River
Trib of Awbeg River
Trib of Assaroola River
Trib of River Loobagh
Trib of Keale River
Trib of River Ogeen
Trib of River Ogeen
Trib of Keale River
Trib of River Loobagh
Barranahown River
Delour River
Trib of River Barrow
Trib of Delour River
Trib of Delour River
Trib of Delour River
Delour River
Delour River
Invertebrate
Site Code
SRAG1
GLASH2
ANNA3
KNICK1
SLAN1
CLOG1
CLOG2
Water Chemistry Site
Code
DWW23(f)
DWW26
DWW4
DWW6
DWW7
DWW8
DWW9
County
Wicklow
Wicklow
Wicklow
Wicklow
Wicklow
Wicklow
Wicklow
Grid
Reference
O 097 135
O 065 016
O 065 033
S 998 952
S 995 937
O 130 074
O 126 076
Easting
309741
306500
306512
299854
299593
313029
312627
Northing
213564
201700
203325
195214
193771
207418
207642
G1
G11
G1
G11
Galway
Galway
R 627 996
M 548 106
162760
154844
199671
210646
G12
G12
Galway
M 565 052
156586
205222
G13
G15
G16
G18
G13
G15
G16
G18
Galway
Galway
Galway
Galway
M 565 052
L 830 498
L 827 493
L 740 513
156556
83047
82718
74043
205208
249800
249356
251332
G2
G3
G4
G5
G6
G7
G8
G9
K1
K2
K3
L10
L12
L13
L14
L15
L16
L17
L2
L3
L4
L5
L6
L8
L9
LS1
LS10
LS12
LS13
LS14
LS2
LS3
G2
G3
G4
G5
G6
G7
G8
G9
K1
K2
K3
L10
L12
L13
L14
L15
L16
L17
L2
L3
L4
L5
L6
L8
L9
LS1
LS10
LS12
LS13
LS14
LS2
LS3
Galway
Galway
Galway
Galway
Galway
Galway
Galway
Galway
Kerry
Kerry
Kerry
Longford
Longford
Longford
Longford
Longford
Longford
Longford
Longford
Longford
Longford
Longford
Longford
Longford
Longford
Laois
Laois
Laois
Laois
Laois
Laois
Laois
M 566 015
M 612 065
M 563 105
M 160 363
M 112 350
M 085 327
M 099 328
M 086 359
Q 969 170
Q 982 182
Q 990 204
R 702 244
R 721 237
R 618 188
R 616 187
R 597 190
R 597 190
R 817 227
R 638 204
R 650 186
R 638 168
R 646 171
R 660 175
R 694 217
R 684 242
N 285 029
N 331 054
S 246 970
S 237 966
S 225 967
N 281 031
N 285 032
156641
161215
156391
116027
111216
108534
109902
108638
96915
98243
99014
170264
172168
161896
161672
159776
159757
181764
163838
165078
163859
164626
166024
169474
168480
228597
233181
224699
223754
222549
228188
228549
201559
206594
210551
236399
235073
232725
232848
235904
117058
118241
120483
124474
123782
118844
118760
119073
119065
122760
120431
118651
116884
117113
117505
121787
124258
202935
205441
197009
196687
196740
203180
203287
80
Main system
Delour River
Delour River
Delour River
Trib of Mountrath River
Trib of River Barrow
Trib of Glenummera River
Trib of Glenora River
Glennafrankagh
Trib of Altderg River
Coolin River
Sruffaunmuinganierin
Trib of Glenummera River
*
Trib of Glencullin River
Trib of Glencullin River
*
Trib of Glenamoy River
Trib of Glenamoy River
Bellanaminnaun River
Trib of Sheep River
Trib of River Aherlow
Trib of River Aherlow
Trib of River Aherlow
Trib of River Aherlow
Trib of River Aherlow
Trib of River Aherlow
Trib of River Aherlow
Clydagh River
Clydagh River
Trib of River Aherlow
Trib of Sheep River
Trib of River Aherlow
Trib of Burncourt River
Trib of Burncourt River
Trib of Burncourt River
Trib of Burncourt River
*
Trib of River Suir
Trib of River Aherlow
Trib of Licky River
Trib of Licky River
Trib of Licky River
Trib of Licky River
Trib of Goish River
Trib of Goish River
Goish River
Invertebrate
Site Code
LS4
LS5
LS6
LS7
LS8
M1
M10
M11
M12
M13
M14
M2
M3
M4
M5
M6
M7
M8
M9
T1
T10
T11
T12
T13
T14
T15
T16
T17
T18
T19
T2
T20
T3
T4
T5
T6
T7
T8
T9
W1
W2
W4
W6
W7
W8
W9
Water Chemistry Site
Code
LS4
LS5
LS6
LS7
LS8
M1
M10
M11
M12
M13
M14
M2
M3
M4
M5
M6
M7
M8
M9
T1
T10
T11
T12
T13
T14
T15
T16
T17
T18
T19
T2
T20
T3
T4
T5
T6
T7
T8
T9
W1
W2
W4
W6
W7
W8
W9
County
Laois
Laois
Laois
Laois
Laois
Mayo
Mayo
Mayo
Mayo
Mayo
Mayo
Mayo
Mayo
Mayo
Mayo
Mayo
Mayo
Mayo
Mayo
Tipperary
Tipperary
Tipperary
Tipperary
Tipperary
Tipperary
Tipperary
Tipperary
Tipperary
Tipperary
Tipperary
Tipperary
Tipperary
Tipperary
Tipperary
Tipperary
Tipperary
Tipperary
Tipperary
Tipperary
Wexford
Wexford
Wexford
Wexford
Wexford
Wexford
Wexford
Grid
Reference
N 276 038
N 297 033
N 295 019
S 348 997
N 366 079
L 905 676
G 046 339
G 028 347
G 010 322
G 047 283
G 072 288
L 896 674
F 855 178
F 911 255
F 898 261
F 878 309
F 909 331
F 951 358
G 066 365
R 893 204
S 016 279
S 002 280
R 993 279
R 988 280
R 979 284
R 948 263
R 945 263
R 887 255
R 884 253
R 901 279
R 906 201
R 891 279
R 907 220
R 908 220
R 922 205
R 955 200
R 991 247
S 029 270
S 017 279
X 212 855
X 224 878
X 174 848
X 190 859
X 170 885
X 159 904
X 171 894
Easting
227675
229712
229594
234839
236680
90547
104608
102811
101088
104744
107280
089611
085545
091155
089800
087822
090976
95114
106695
189359
201647
200240
199374
198875
197987
194857
194527
188710
188495
190180
190631
189140
190702
190815
192276
195582
199159
202978
201719
221260
222457
217494
219066
217047
215946
217105
Northing
203807
203334
201996
199782
207933
267687
333981
334720
332253
328324
328874
267419
317826
325560
326190
330950
333125
335885
336507
120401
127960
128048
127905
128086
128490
126398
126387
125536
125348
127915
120157
127993
122059
122056
120535
120028
124754
127038
127950
85579
87880
84858
85973
88586
90467
89485
81