aessaassa•••••••••••••••••••••••••case flows) than...
TRANSCRIPT
aessaassa•••••••••••••••••••••••••
••••••••••••••••••••••••••••••••••
Ecologically Acceptable Flows Phase II
Volume 1: Habitat Time Series Analysis
R&D Technical Report W19
M.J.Dunbar, C.R.N.Elliott, M.C.Acreman & A.Gustard
Research Contractor:Institute of Hydrology
This report is an official document preparedunder contract between the Institute ofHydrology and the Environment Agency. Itshould not be quoted without permission of theInstitute of Hydrology and the EnvironmentAgency
R&D Technical Report W19
Publishing OrganisationEnvironment AgencyRivers HouseWaterside DriveAztec WestAlmondsburyBristol 8S12 4UD
Tel: 01454 624400 Fax: 01454 624409
@Environment Agency 1996
All rights reserved. No part of this document may be produced, stored ina retrieval system, ortransmitted, in any form or by any means, electronic, mechanical, photocopying, recording orotherwise without the prior permission of the Environment Agency.
The views expressed in this document are not necessarily those of the Environment Agency.Its officers, servants or agents accept no liability whatsoever for any lossor damage arisingfrom the interpretation or use of the information, or reliance upon views contained herein.
Dissemination status
Statement of useThis report forms part of Environment Agency R&D Project 282 Phase II'EcologicallyAcceptable Flows'. It details work carried out by the Institute of Hydrology to combine theoutputs from the habitat modelling procedures undertaken in Phase I of the project, with flowtime series data from the National River Flow Archive (NRFA), to developthe techniques forpredicting habitat availability on a temporal basis.
•Research contractorThis document was produced under R&D project 282 by:
•Instituteof HydrologyWallingfordOxon. OXIO 8BBUK
•Tel: 01491 838800Fax: 01491 692424
Environment Agency's Project ManagerThe Environment Agency's Project Manager for R&D Project 282 was:Dr Ten-yNewman - EA South Western Region
•
•••
Internal: Released to RegionsExternal: Released to Public Domain
EXECUTIVE SUMMARYThis report forms part of Environment Agency R&D Project 282 Phase Il 'EcologicallyAcceptable Flows'. It details work carried out by the Institute of Hydrology to combine theoutputs from the habitat modelling procedures undertaken in Phase I of the project, with flowtime series data from the National River Flow Archive (NRFA), to develop the techniques forpredicting habitat availability on a temporal basis. In the Phase I project, the Physical HabitatSimulation (PHABSIM) system, a computer model of instream physical habitat developed inthe United States and applied world-wide, was investigated for use in the UK. PHABS1Mhydraulic models for nine British rivers wcre produced. Then, using a habitat model, thehydraulic output was combined with habitat suitability indices for fish (brown trout Salmotrutta, roach Rutilus rutilus and dace Leucisus cephalus), plus suitable invertebrate andmacrophyte groups to produce theoretical relationships between physical habitat area (WUA)and river discharge (Q).
It is logical to assume that future population levels in an instream aquatic community will be influenced not only by physical habitat at that future time, but also by the patterns of physical habitat leading up to it. Thus extending the 'traditional' PHABSIM model results (theWeighted Usable Area versus Discharge curve) to temporal predictions of habitat is a crucialstep in relating model output to changes in fish and invertebrate populations. In this study, theapplication of several existing low-flow analysis techniques to the temporal analysis ofinstream physical habitat have been investigated. Calibrated PHABSIM habitat models werecombined with flow time series data from the National River Flow Archive (NRFA), in orderto predict habitat availability on a temporal basis. Where possible, comparisons have beenmade between results from different rivers.
•Naturally, the flow regime of a river will be an important factor in determining temporalpatterns of habitat. In this study, the greater flow variability in upland catchments (comparedto less flashy lowland sites) was reflected in higher habitat variability. However, usingcurrently available habitat suitability indices, the non-linear relationship between habitat anddischarge also has profound influence on the form of a habitat time series. This additionalsource of variation has been shown to be especially important for fish life stages, whichgenerally show optimum habitat at intermediate flows.
•A fundamental method for analysis of time series of river flows is to derive a cumulativefrequency diagram, this is often known as the flow duration curve. Following this concept, habitat duration curve analysis was undertaken for the nine UK rivers. This report discusses the methods used to create the curves and further analysis that can aid in their interpretation.These methods are of particular use in the analysis of how alternative flow regimes affecthabitat available to individual life stages of a species. Techniques for aggregation of dailyhabitat values to monthly and annual statistics are discussed, as is the use of single - yearhabitat duration curves.
•
The study discusses how variations in the flow regime and the habitat - discharge relationshipaffect the shape of a habitat duration curve. Plotting corresponding flow values on a habitatduration curve can aid in its interpretation, particularly when intermediate habitat values mayarise from both high and low flows. For this type of flow - habitat relationship, numbers offlows outside the model calibration, combined with relative habitat values at the model limits,
•EA Technical Report W19
can have a significant influence on the shape of the habitat duration curve. This emphasisesthe importance of a robust PHABSIM calibration, using the widest possible flow range.
Application of another low-flow technique, flow-spell analysis is also discussed, as aremodifications to aid in habitat time-series interpretation. Habitat-spell analysis characterisesthe lengths of time the habitat drops below a certain threshold value, andit has been suggestedthat analysis of these deficit periods could provide further insight into possible critical habitatlimitations. Further development of this promising technique is required. This report alsobriefly reviews options for further research into the sensitivity of habitat time series tovariations both in habitat suitability data, and in the transfer of gauged flow time-series toungauged study sites.
EA Technical Report W19 iii
••••••••••••••••••••0
•••
Contents
Executive Summary
Introduction1.1Background
1.2IFIM and PHABS1M rationale and concepts1.3Analysis of flow and habitat time series
1.3.1Habitat time series1.3.2Summary habitat statistics1.3.3Habitat duration curves1.3.4Points to note in duration curve analysis of habitat data1.3.5Flow and habitat spell analysis
2Review of PHABSIM output from Ecologically Acceptable Flows Phase 12.1Background
2.2Missing values and out of range flows
3Methodology
3.1Time series analysis3.2Duration curve analysis3.3Combination of flow data with habitat duration curves3.4Single-year duration curves3.5Flow and habitat spell analysis
4 Results
4.1River Exe at Warren Farm4.2River Wye at Pant Mawr (including comparison with Exe)4.3River Hodder at Hodder Bank4.4River Blithe at Hamstall Ridware4.5River 1tchen a's of Highbridge4.6River Lymington u/s of Balmer Lawn4.7Millstream at WE East Stoke4.8River Lambourn at Hunt's Green (including comparison with 1tchen)4.9River Gwash at Belmesthorpe4.10River Lymington: habitat duration curves combined with flow data4.11River Gwash: plotting missing values on time scries and duration curves4.12Lamboum and Wye: comparison of single-year duration curves4.13Lambourn and Wye: flow and habitat spell analysis
5Conclusions and Recommendations
References
A ppendicesASummary habitat statistics
Relationships between Weighted Usable Area (WUA) and Discharge (Q)Habitat time series and duration curvesFurther analysis: missing values, adding flow to a habitat duration curve; single-yearduration curves, flow & habitat spell analysis
Page
1
2
4
4
5
56
7
8
8
10
12
13
1314
15 15
17
17
1819
1920
20
21
21
23
23
24
25
32
34
43
90
••
EA Technical Report W19 iv•
List of tables
2.1 Nine PHABS1M sites calibration flows; model ranges; out of rangeand missing flows 8
3.1 Summary of gauging stations close to study sitcs II3.2 Output for combined habitat / flow duration curve 13
4.1 Lambourn and Wye: variation in 95 percentile flow and habitat 23
A 1 Summary of gauged flow and habitat statistics 30
EA Technical Report W19
110
List of figures
1.1 PHABSIM model 3I 1.2 Development of a habitat time series 41.3 Habitat duration curve 6D 1.4 Flow & habitat spell analysis 6
IP 2.1 Location of study sites 9
3.1 Summary of time periods of gauging station data 10
ID 4.1 Rivers Exc and Wye: comparison of flow frequencies and habitat 164.2 River Blithe: adult dace habitat 1811, 4.3 River Itchen: habitat time series for spawning brown trout, Nov and Dec 1970-73 194.4 River Lambourn: habitat duration curve for adult brown trout 204.5 River Gwash: WUA v Q curve for adult brown trout 22
5.1 Variations in Weighted Usable Area v Discharge relationship 27
B1 River Exe at Warren Farm: habitat vs discharge relationships 3582 River Wye at Pant Mawr: habitat vs discharge relationships 36B3 River Blithe at Hamstall Ridware: habitat vs discharge relationships 37B4 River lichen upstream of Highbridge: habitat vs discharge relationships 38B5 River Lymington u/s of Balmer Lawn: habitat vs discharge relationships 39B6 Millstream at WE East Stoke: habitat vs discharge relationships 40
0 B7 River Lambourn at Hunt's Green: habitat vs discharge relationships 4188 River Gwash at Belmesthorpe: habitat vs discharge relationships 4241CI.1 River Exe: time series for flow, total habitat, adult and juvenile brown trout 44C1.2 River Exe: time series for fry and spawning brown trout 45CI.3 River Exe: time series for Chlorperlidae (A), Leuctridae (A) 46C1.4 River Exe: duration curves for flow, total habitat, all life stages of brown trout 47C1.5 River Exe: duration curves for Chlorperlidae (A), Leuctridae (A) 48C2A River Wye: time series for flow, total habitat, adult and juvenile brown trout 49C2.2 River Wye: time series for fry and spawning brown trout 50C2.3 River Wye: time series for Lcuctridae (A), Polycentropusflavomaculatus, Ranunculus 51C24 River Wye: duration curves for flow, total habitat, all life stages of brown trout 52C2.5 River Wye: duration curves for Leuctridae (A), Rolycentropus flavomaculatus,
Ranunculus 53C3.1 River Blithe: time series for flow, total habitat, adult and juvenile dace 54C3.2 River Blithe: time scries for fry and spawning dace 55C3.3 River Blithe: time series for flow, total habitat, adult and juvenile roach 56C3.4 River Blithe: time series for fry and spawning roach 57C3.5 River Blithe: time series for Gammaridac (A), Leuctridae (A) 58C3.6 River Blithe: duration curves for flow, total habitat, all life stages of dace 59C3.7 River Blithe: duration curves for flow, total habitat, all life stages of roach 60 C3.8 River Blithe: duration curves for Leuctridae (A) Rolycentropus flavomaculatus,
Nasturtium61
C4.1 River ltchen: time series for flow, total habitat, adult and fry / juvenile brown trout 62C4.2 River lichen: time series for spawning brown trout 63C4.3 River lichen: time series for Gammarus pulex, Ephemeridae(A) 64C4.4 River Itchen: duration curves for flow, total habitat, all life stages of brown trout 65 C4.5 River Itchen: duration curves for Gammarus puler, Ephemeridae (A), Ranunculus,
Nasturtium66
C5A River Lymington: time series for flow, total habitat, adult and juvenile brown trout 67C5.2 River Lymington: time series for fry, spawning brown trout 68C5.3 River Lymington: time series for Gammaridae (A), Leuctridac (A) 690
EA Technical Report WI9 vi•
II
5
ID
ID
•
•
0
••
•
•
•
•
•
•
•
•
•
•
•
C5.4 C5.5
C6.1C6.2C6.3C6.4C6.5C6.6C6.7C6.8C7.1C7.2C7.3C7.4C7.4
C8.IC8.2C8.3C8.4C8.5
DlaDI bD2aD2bD3aD3b
D4D5
i 1%dD7ID8./09
-/ DI 0/D11
I DI2i DI3
River Lymington: duration curves for flow, total habitat, all life stagesof brown troutRiver Lymington: duration curves for Gammaridae (A), Leuctridae (A), Ranunculus.NasturtiumMillstream: time series for flow, total habitat, adult and juvenile daceMillstream: time series for fry, spawning dace,Millstream: time series for flow, total habitat, adult and juvenile roachMillstream: time series for fry, spawning roachMillstream: time series for Gammaridae (A), Leuctridae (A)Millstream: duration curves for flow, total habitat, all life stagesof daceMillstream: duration curves for flow, total habitat, all life stagesof roachMillstream: duration curves for Gammaridae (A), Leuctridae (A), RanunculusRiver Lambourn: time series for flow, total habitat, adult and fry/juvenile brown troutRiver Lamboum: time series for spawning brown troutRiver Lamboum: time series for Gammaridae (A), Ryacophilia dorsalisRiver Lambourn: duration curves for flow, total habitat, all life stagesof brown troutRiver Lambourn: duration curves for Gammaridae (A), Ryacophilia dorsalis,Ranunculus, NasturtiumRiver Gwash: time series for flow, total habitat, adult andjuvenile brown troutRiver Gwash: time series for fry, spawning brown troutRiver Gwash: time series for Gammaridae (A), Sericostomatidae (A)River Gwash: duration curves for flow, total habitat, all life stagesof brown troutRiver Gwash: duration curves for Gammaridac (A), Sericostornatidae (A), Nasturtium
River Lymington: habitat duration curve with flows for brown trout fryRiver Lymington: flow duration curve with habitat for brown trout fryRiver Lymington: habitat duration curve with flows for NasturtiumRiver Lymington: flow duration curve with habitat for NasturtiumRiver Lymington: habitat duration curve with flows for RanunculusRiver Lymington: flow duration curve with habitat for Ranunculus
River Gwash: time series for adult trout with out-of range values set to model limitsRiver Gwash: duration curve for adult trout with and without out-of range valuesset tomodel limitsRiver Wye: yearly flow duration curves (1966-1977)River Wye: yearly habitat duration curves for adult brown trout (1966-1977)River Wye: yearly habitat duration curves for fry brown trout (1966-1977)River Larnbourn: yearly flow duration curves (1966-1977)River Lambourn: yearly habitat duration curves for adult brown trout (1966-1977)River Lambourn: yearly habitat duration curves for fry / juvenile brown trout (1966-1977)River Wye: flow & habitat spell analysis, adult, juvenile and fry brown trout.River Lambourn: flow & habitat spell analysis, adult and fryrjuvenile brown trout
70
71727374757677787980818283
848586878889
919293949596
97
9899
100 100101102 102103104
•Abbreviations usedIH Institute of HydrologyNBS US National Biological ServiceIFE Institute of Freshwater Ecology, WarehamMF mean flowNWA National Water ArchiveNRFA National River Flow ArchiveQ River discharge (mcasurcd in cumecs)
Other conventionsHabitat suitability curves developed for invertebrates in Project 132.1Phase I was taken from the RIVPACS
ID database. Species names denote criteria developed from data on occurrence of that species,while generic names (e.g. Leuctridae (A)) refer to curves developed with data on abundance of that genus, aswell as occurrence.•EA Technical Report W19 vii•
AcknowledgementsWe would like to thank the staff of the Environment Agency and the Institute of FreshwaterEcology for habitat suitability data and advice. We are also grateful to the staff of theEnvironment Agency for providing river flow data. We acknowledge the US NationalBiological Service Instream Flow Group, for the development of PHABSIM and for theirassistance in applying the model in the UK. The authors would also like to thank theircolleagues at IH for their assistance including Andy Young, Ann Sekulin and Sam Green.
•
••••
•
EA Technical Report W19 viii•
I•
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
•••
1.INTRODUCTION•
1.1BackgroundFollowing concern over low flow conditions arising from recent droughts in the UK, anational low flows assessment by the NRA (1993) identified a priority list of 40 locationsperceived as suffering from excessive abstraction. This, coupled with the requirementunder 1989 Water Act for the NRA to set Minimum Acceptable Flows when requested bythe Secretary of State, has prompted the need to develop operational tools for managingaquatic communities in British rivers on a national scale.
•A previous NRA R&D project (Johnson et at 1993a) completed by the Institute ofHydrology, investigated the potential for the use in the UK of PHABSIM, a computermodel of instream physical habitat, developed in the United States. This model can beused to predict the reduction in habitat available to various target species resulting fromlow river flows.
• In addition to its routine application in the United States, PHABSIM is the subject ofprevious studies and ongoing research in several countries world-wide including Canada(Shirvell & Morantz 1983), New Zealand (Scott & Shirvell 1987), Norway (Heggenes1990) and France (Souchon et al. 1989). In the UK, under commissions from MAFF(Johnson et al, 1993c, Elliott et at 1995b), PHABSIM has also been used in theenvironmental assessment of two flood defence schemes in the Thames catchment.
•In NRA South Western Region, the Institute of Hydrology has applied PHABSIM tosalmonid habitat at two sites on the river Allen (to assess the impact of the historicalgroundwater abstraction regime which has reduced river flows) (Johnson et aL 1995c),the rivers Bray and Barle (Johnson et aL 1994b), and the river Piddle (Johnson and Elliott19956). An initial investigation into the use of PHABSIM to predict juvenile cyprinidhabitat on the Thames, has recently been completed (Elliott et at 1995), and WorcesterCollege of Further Education is currently working on two PHABSIM studies on the riversKennett and Tavy, the latter in collaboration with IH.
•
For this project the Environment Agency has commissioned the Institute of Hydrology tocombine the calibrated habitat models from the Phase I R&D project (Johnson et al.1993a) with flow time series data from the National River Flow Archive (NRFA), inorder to predict habitat availability on a temporal basis. In particular, the use of thehabitat duration curve to predict impacts across the range of flows experienced in a riverhas been examined. Extending the PHABSIM model to temporal predictions of habitat isa crucial step in relating model output to fish and invertebrate populations.
•This report also investigates the application of several other existing low-flow analysistechniques to the temporal analysis of instream physical habitat, and provides suggestionsas to how these techniques could be further improved for habitat analysis.
•••
EA Technical Report W19 1•
1.2 IFIM and PHABSIM rationale and conceptsThe Physical HABitat SIMulation system (PHABSIM) is part of a wider framework, theInstream Flow Incremental Methodology (IFIM), developed ai the US Fish and WildlifeService by an interdisciplinary team of scientists (Stalnaker 1993, Bovee 1995).
IFIM is a framework for analysing spatial and temporal changes in habitat characteristicssuch as streamflow and channel geometry. Its most commonly-used component isPHABSIM, but it also includes models for water quality, water temperature or indeed anyother model which simulates characteristic features which could influence habitat. It alsoincludes mechanisms for analysing the institutional aspects of water resource issues, plusthe all-important scoping and planning process, along withtechniques for negotiation andresolution.
PHABSIM is a computer model running under the MS-DOS operating system thatprovides an estimate of physical habitat loss or gain resulting from changes in discharge.It is calibrated by making accurate field survey measurements of channel geometry attransect sites on a river system, along with measurements of watcr surface level at two ormore flows, and stream velocities at onc or more flows (Trihey & Wegner 1981, Bovee1982, Johnson et at 1991). Output from PHABSIM is in the form of a Weighted UsableArca (WUA) versus Discharge (Q) curve, for each species / life stage of interest. Thisgraph shows how habitat (represented by WUA) varies with streamflow. Thcre has beenmuch debate as to the interpretation of these curves (Orth 1987, Mathur a al. 1985, Scott& Shirvell 1987, Gore & Nestler 1988), however it is certain that one cannot simplychoose an ecologically acceptable minimum by choosing the flow corresponding tomaximum available habitat. Further analysis is required. One such approach, as outlinedhere, is to carry out time series analysis on the predictions of available habitat within ariver.
The underlying concepts of PHABSIM are:
It is habitat-based, with potential usable habitat being simulated for unobservedflow or channel conditions.
Target species exhibit a quantifiable preference/avoidance behaviour to one ormore of the physical microhabitat variables; velocity,depth, cover or substrate.
Prcferred conditions can be represented by a suitability index which has beendeveloped in an unbiased manner.
Individuals select the most preferred conditions within a stream, but will use lessfavourable areas with decreasing frequency/preference.
Species populations respond to changes in environmental conditions thatconstitute habitat for the species.
The purpose of the PHABS1M system is the simulation of the relationship betweenstreamflow and available physical habitat where physical habitat is defined by themicrohabitat variables depth, velocity and substrate/cover. The two basic components of
EA Technical Report W19 2
PHABSIM arc the hydraulic and habitat simulations within a stream reachusing definedhydraulic parameters and habitat suitability criteria, as displayed in Figure 1.1 below.
• XSEC2 WSL21
•
•
XSEC 1
WSL1
o-
c73
o-5
•
•
CHANNEL INDEX ANDSIMULATION DEPTHS
AND VELOCITY100
7\VELOCITY DEPTH SUBSTRATE
•• 3
0DISCHARGE
Figure 1.1 PHABS1M
The PHABSIM hydraulic simulation model is calibrated with field survey data and isable to model depths and velocities (along with substrate and / or cover as unchangingparameters), at calibration and non-calibration flows. The PHABSIM habitat model thenreceives these data and combines them with habitat suitability indices for target species /life stages. For each target life stage, it produces a single index, the Weighted UsableArea (WUA) at each modelled flow.
It is clear that in conducting a PHABS1Mstudy, an ideal goal would be to relate changesin aquatic populations to change in the flow regime. Although some studies havesuccessfully demonstrated that PHABSIM may be capable of achieving this goal, it mustbe appreciated that PHABSIM alone is not capable of this task since it predictschange ina weighted measure of physical habitat area (WUA) available to aquatic species and docsnot predict change in biomass. In some instances a linear relationship between biomassand WUA has been demonstrated (Milhous, 1988, Jowett, 1992) but it is clear that over aperiod of years, this will not be the case in most rivers since for some of thetime, factorsother than physical habitat may be limiting to populations. However in the analysisof the impacts of low flows, lack of physical habitat will clearly be a key factor.
In the absence of equivalent population models, one accepts the limitation of using WUAas the key variable and attempts to take into account, as rigorously as possible, factorswhich are likely to influence the relationship between WUA and populations. Gore andNestler (1988) make the following statement with regard to this issue:
•EA Technical Report W19 3
"PHABSIM is a vehicle for presenting biological information in a format suitable forentry into the water resources planning process. It is not nor was it ever intended to be,a replacement for population studies, a replacement for basicreseurch into the subtletiesoffish or benthic ecology, nor a replacement for biological innovation or common sense.As such, PHABSIM has been found to be a defensible technique for adjudicating flowreservations".
1.3 Analysis of flow and habitat time series
1.3.1. Habitat time seriesA time series is a sequence of events, arranged in order of occurrence. In this study,historical time series of flows taken from gauging station data held either by the NationalRiver Flow Archive (NRFA) at the Institute of Hydrology, or regional EA offices, werecombined with PHABSIM habitat model output to predict the temporal patterns of habitatthat would have occurred during the period of record.
The major premise of habitat time series analysis is that habitat is a function ofstreamflow:
Habitat(t) = f(Q(t))
where t=time
Time series of available habitat (as represented by WUA)may be used to analyse theimpacts on river biota of alternate flow regimes; this is a major advantage of PHABSIMover multi-variate statistical models. It provides the first stage in linking the 'traditional'PHABSIM output - the WUA v Discharge curve, with the recognition that single andmultiple critical events can limit physical habitat, and thushave major impacts on riverbiota. A more complete assessment of- the effects of a hydrological regime requiresconsideration of all life stages, and where required, targeting the analysis to crucial timeperiods. For example, for fish species, critical flows during spawning or fry developmentwill affect recruitment and thus have a knock-on effect on the adult population, even ifadult habitat is good in following years.
EA Technical Report WI9 4
Figure 1.2. Development of a habitat time series
1.3.2. Summary habitat statisticsOnce a daily habitat time series has been calculated, the user may wish toaggregate dailyvalues to monthly or even yearly, in order to examine habitat changes on a long-termbasis. Commonly-used summary statistics include (Milhous 1990b):
Mean habitatMedian habitatIndex-A: mean of all the habitats between 50% and 90% exceedance, i.e. the majorityof the low flow eventsIndex-B: mean of all habitats between 10% and 90%Minimum habitat : not recommended in most casesMaximum / optimum habitatAn exceedance statistic e.g. 90, 95 percentile habitat (see section 4.12 for example)Number of days below threshold, or total threshold deficit (see section 4.13 forexample)
1.3.3. Habitat duration curvesHabitat time series may be subjected to further analysis using the techniques developedfor river flow analysis. The first example of this is the habitat duration curve (Milhous1986, Gordon McMahon and Finlayson 1992, Vogel and Fennessey 1995). A durationcurve, whether for flow, habitat or another instream variable, displays the relationshipbetween the variable and the percentage of time it is exceeded.
Duration curves are constructed by sorting the data (time series of flows or habitat values)from highest to lowest, and expressing each data point as a percentage of the total numberof values.
EA Technical Report W19 5
3
02 3 4
Discharge (=secs)
5 10 2D 30 50 7000 90 95% of 5me exceeded
'rave
Time
!PS.
Time
I I.
When a flow-duration curve from UK catchments are plotted with a normal probability(x) axis and logarithmic flow (y) axis, the curve approximates to a straight line indicatingthat the logarithm of discharge follows a normal distribution. However this is less oftenthe case for habitat duration curves.
Duration curves arc particularly useful for assessing the impacts of alternative flowregimes. Johnson et aL (1993b, 1994b, 1995b) demonstrated their use on several rivers inthe EA South Western region.
Left portion of curve,optimum habitat
5 10 20 30 50 70 80 90 95% of time e ceededRight portion of curve,
infrequent low habitatCentral portion of curve.events (in this case intermediate habitat arising from caused by lowest flows)flows both lower and higher thanoptimum
Figure 1.3. Example of a habitat duration curve
1.3.4. Points to note in duration curve analysis of habitat dataIt is important to note that a habitat duration curve shows exceedance percentiles forvalues of habitat, ranked from highest to lowest. It does notsimply show flow exceedancepercentiles mapped to their corresponding habitat values. It is also important to note thatespecially for fish life stages, there is not generally a continuous increasing relationshipbetween habitat and flow. Instead, habitat often increases with flow up to a certain flow,and then decreases. There are many examples of this in the results of R&D project B2.l.The results of this phenomenon are that some central portion of the habitat duration curvewill be made up of habitat arising from both high and low flows. This is unlike a flowduration curve, where high flows are always on the left and low on the right. Instead,'optimum habitat' arising from intermediate flows forms the left hand side of the curve(see Figure 1.3), combinations of intermediate habitat arising from either higher or lowerthan optimum flows comprise the central portion, and depending on the range of thecalibrated model and range of flows in the time series, the right hand portion willcomprise of low habitat figurcs arising from either the highest or lowest flows.
While not detracting from the use of habitat duration curves in the evaluation of theimpacts of alternate flow regimes, caution should be exercised in their interpretation. It ispossible to plot values of flow on a habitat duration- curve-to aid in interpretation, thisprocedure is further discussed in section 4.6 below.
30
20
I0
5
3
2
50!
: Summer Winter
WUA
(%
habitat
available)
EA Technical Report W19 6
In this study, habitat-discharge curves for invertebrates and macrophytes exhibit eithercontinuous increasing or continuous decreasing behaviour. This one-to-one mapping offlow and habitat makes interpretation of their habitat duration curves morestraightforward.
Duration curves also do not give information about the temporal distribution of low flow or habitat events, i.e. whether they occur together or separately. Other techniques such asflow & habitat spell analysis have been developed for this purpose. •
•1.3.5. Flow and habitat spell analysisThis technique, described in the Low Flow Studies Report (IH, 1980) characterises periods of flow below a certain threshold value. Its advantage over the flow durationcurve is that it considers the temporal distribution of low flow events. This technique hasbeen applied to simulated habitat (Capra, Breil and Souchon, 1995). The methodology isdiscussed in Section 3 below.
•
•••••••••
100
3-:,,...I tou ITO
ft •is 8 10
sO
io
to= • .•
i 41' ' •-•....W
0 2 4 0 a 10 12 14
Time CurmAMin mantas Ont.ana(%Ot IOW ttudred datial)
•Figure1.4.Flow/habitat veil analysis.
120
EA Technical Report WI9 7
2. REVIEW OF PHABSIM OUTPUT FROM PHASE I OFTHE ECOLOGICALLY ACCEPTABLE FLOWS PROJECT
2.1. BackgroundDuring NRA R&D Project EAF Phase I, calibration datawere collected from ten sites,each representing one ecological group defined using data from the RIVPACS database(Wright et at 1988, Johnson et at 1993a). It was not possibleto calibrate the PHABSIMmodel for the Great Ouse / Lees Brook, so this site is not considered further iii this report.
The following table indicates the calibration flows (CI, C2, C3) for the remaining studysites along with the lower and upper limits of the model (MLowand MHigh). It also liststhe out-of-range flows for summer and winter separately, when the time series of dailymean flows was run through the model. The time series were 16 years (5844 days) exceptwhere stated.
Table 2.1. Table of calibration flows, model ranges and out-of-range flows for EAFPhase I sites
Site CI
(m3/s)
C2 C3 MLow MHigh no. flows below lower
limit
no. flows above
upper limit
no. flowsmissing
S W
SW S WExcatWarren 0.87 0.21 0.12 0a57 1.70 359 3 0 13 0 0Earm
(Usingtimeseriestransferredfrom
Bark at Ilrushford)
Wyc at Pant Mawr 0.55 1.87 263 a 14 4.53 203 23 128 352 0 33Hodder at Hodder 2.24 2.83 5 27 0.42 12.74
Bank
BlitheatHamstall 0.46 0.59 0.48 0.142 2.83 0 0 63 301 344 247Ridwarclichenu/sof 3.41 5.1
0.42 7.64 0 0 III 560 0 0Ilighbridge
Lymingtonat 0.38 0.09 0.59 0.057 2 69 203 0 102 459 42 0Balmer Lawn
Millstream at IFE 0.52 1.77 2.01 0.11 4.53 189 60 0 0 0 0East Stoke
Lambourn at Hunt's 0.55 0.68 0.78? 0.42 5.66 121 41 0 0 0 0Green
Gwashat 0.37 1.17 1.35 014 1.70 0 0 132 385 0 0Belmesthorpe
s=summer, w=winter
2.2. Influence of missing values and out of range flows on time seriesand duration curvesOf particular note are the numbers of flows above the modelrange for almost all sites (allexcept the Lambourn and Millstream). Given the time that would be required to re-calibrate the models to obtain more complete calibrations, and as this report is mostlyconcerned with the analysis of low flow events, this was notconsidered a major problem.
Currently, the IH PHABSIM time series analysis software simply ignores missing values.In contrast, the flow-spell analysis software developed for -IH (1980) and Gustard,Bullock and Dixon (1992) interpolates between the known values that bound a region of
-
EA Technical Report W19 8
missing data. The influence of the missing data on the form of habitat duration curves isinvestigated below.
I. Exe at Pixton
Wye at Pant Mawr
•4. Blithe at Hamstall Ridware
Hodder at Hodder Bank
• 5. Itchen at Highbridge and AllbrookLymington at Balmer LawnMillstream at IFE East StokeLambourn at Hunt's Green
Gwash at Belmesthorpc•••
3
sc>4
2 •9
1
Figure 2.1 Location of study sites
EA Technical Report W19
8 5
6
7
9
3. METHODOLOGY
3.1. Time series analysisUnlike the wealth of literature on procedures for other parts of the PHABS1M process(e.g. collection of suitability data, hydraulic modelling), information on time-seriesprocedures is limited, whether in published papers (Capra et at 1995), the US-specific
'how-to' manual (Milhous 1990b), or UK applied studies from IH (Johnson et oL 1993b,
1994b, 1995b).
In Project B2.I, a computer data file containing the relationship between WUA and Q foreach species/life stage was produced for each study site by running hydraulic simulationoutput through the habitat simulation software. The IH PHABSIM software suite includesa program, WUAQT, which combines a flow time series with the WUA v Q file (known
as a ZHAQFM file), to produce a time series of flow, habitat for all life stages of aspecies, and total habitat. The latter is effectively a measure of total river area per 1000mlength of river. The software can currently analyse daily and monthly mean flow data.WUAQT performs an interpolation on the data in the ZHAQFM file in order to matchany flow within the model range to a habitat value.
The bulk of the analysis was performed on datasets of daily mean flows for each studysite of 16 years duration (1966-1981 inclusive). This period was chosen *asthe longestavailable that covered all required gauging stations, although later changes to proceduresfor the Exe and Hodder required use of different time periods for those two sites. Studysites for NRA R&D EAF Phase I were chosen to be close to NRFA stations, and six ofthc nine sites (Wye at Pant Mawr, Blithe at Hamstall Ridware, lichen U/S of Highbridge,Lymington U/S of Balmer Lawn, Frome at I.F.E. East Stoke(the Millstream), and Gwashat Belmesthorpe) were within 1.5km of the station. The others were situated as follows:
Site Distance from stationR. Exe at Warren Farm - 15 kmR. Hodder at Hodder Bank 12 kmR. Lamboum at Hunt's Green 4 km
ErWPtcton
WyelP ant M/VeT
BlItheM RI*. I-
ttchenMighbrklga
LymIngtonfEtatner Law r
MastreanVIFE Eag IF
LamtournWetfordIf
Gwasheelmesthape si-1
Barlellkustdad I-I
'4'
I0 50 60 70 80
Figure 3.1 Summary of data availability for each gauging station
90 Year
EA Technical Report W19 10
The Micro Low Flows software package (Young, 1992) was used to estimate time seriesat these sites by correcting the gauged flows by a factor representing the difference incatchment characteristics, i.e. (estimated mean flow at site)/(estimated flow at gaugingstation).
Table 3.1 Summary of gauging stations close to the PHABSIM sites.
Mean Q10flow
4.42 11.0 2.52 0.579
1.65 3.94 0.86 0.166
1.23 2.81 0.586 0.320
5.26 7.68 4.80 2.92
0.99 2.55 0.431 0.049
15km from site.Wimblcball reservoira major influence
<1km from site.Natural
At site. Both gaugeand site influenced byBlithefield reservoir
At site. Currentlyminor artificialinfluences
I.5km &Usof sitc
Q50 Q95 NotesRiver Catchment Station Data
tYPe
Exe Upland impervious 45009 BCsandstone
Wye Upland, peat over 55010 AAslates / shales
llodder Peat moorland overmillstone grit andcarboniferouslimestone
Blithe Sandstone and clay 28002 BC
lichen Chalk, very 42010 AApermeable
Lyinington Impervious, clay, 42003 AAsand and gravel
Millstream Chalk, some sand /gravel / clay
EA station only, data obtained from BlandfordForum office
Lambourn Chalk downland 39031 BA 1.02 1.68 0.882 0.409 4km u/s of sitc
39019 AB 1.69 2.76 1.49 0.749 4km cl/sof sitc belowconfluence
Gwash Clay / limestone 31006 AA 0.79 1.45 0.632 0.289 I .5km dis of site
Source: National River Flows Archive and 1H Report 108
Output was calculated as follows:
Flow: daily flow expressed as a percentage of the mean flow of the dataset
Total habitat:
Habitat for species /life stage
daily values expressed as m'/1000m of river. This is effectively the area of riverwhich would be available to a species that had no restrictions in its preference. It ismeasureof total river area per 1000m of river, regardless of habitat quality.
daily values expressed as percentage of total habitat at that flow. This procedureis analogous to the standardisation of flows above.
By expressing habitat for each species / life stage in this way, one is able not only tocompare relative areas of habitat in rivers of different sizes, but also retain the ability tocompare areas of habitat between different species / life stages in the same river.
EA Technical Report W19 11
Note that this procedure does not make any judgements about biomass of various lifestages, or that the biomass of 1 unit of WUA for one life stage at carrying capacity isequivalent to the biomass of one unit of WUA for another. Bovee (1982) discussesmethods that may be used to deduce this relationship, and thus assess how critical levelsof habitat for each life stage may influence population levelson an ongoing basis.
In order to provide a visual image of how habitat is distributed in space and time, timeseries of daily mean flows, along with total habitat and habitatfor all fish and invertebratespecies / life stages are presented in Appendix C. Although 16 years oltime series datawere used in the analysis, either eleven or three years of dataare presented for illustrativepurposes in Appendix C. The number of years displayed was reduced from eleven tothree for the flashier catchments to maintain legibility.
3.2. Duration curve analysisTime series of habitat were then analysed to indicate the proportions of time each habitatvalue was exceeded (duration curve analysis). This was performed using the HABDCprogram, an IH addition to the US PHABS1M software. During the project, the HABDCsoftware was altered to make it more user-friendly. It now accepts a list of habitat time-series files (usually one per species) from the WUAQT programand also enables durationcurves for summer (April-September), winter (October to March), or any user-definableperiod, to be calculated from an annual time series. Output from HABDC is in the formof a data file indicating habitat and percentage exceedance, and plots which appear on-screen and may be printed.
Clearly it is advisable that analysis of habitat time series of certain life stages be restrictedto the months in which these life stages actually occur. The following time periods wereselected as a compromise between a realistic portrayal of the conditions that would beexperience by the life stage and the desire to keep to the simplest possible analysis.
Brown trout (Salmo trutta)
Adult & Juvenile
Fry / Juvenile (Wessex
curves)
Fry
Spawning
Dace (Leucisus cephalus)
Adult & Juvenile
Fry
Spawning
Roach (Radius mutt's)
Adult & Juvenile
Fry
Spawning
All invertebrates
All macrophytes
Summer and winter
Summer and winter
Summer only
November-December
Summer and winter
Summer only
March-April
Summer and winter
Summer only
April-June
Summer and winter
Summer only
Duration curves all species / life stages are presented in Appendix C, along with curvesfor flow and total habitat. Also presented in Appendix A are 5%, 50% and 95%exceedance percentiles for summer and winter habitat for all fish life stages, and as ameasure of habitat variability, the ratio (5th percentile)/(95thpercentile) habitat.
EA Technical Report W19 12
0
0
03.3. Combination of flow data with habitat duration curvesAs mentioned above, the nature of some WUA v Q relationships - i.e. rising to a peakWUA then falling, can make interpretation of duration curves difficult. An example isgiven in Appendix D (figures DI-D3) of some habitat duration curves from the riverLymington where flow is also plotted on a habitat duration curve. The III duration curvesoftware currently does not allow this, instead a habitat duration curve was constmcted
0 from first principles in a spreadsheet, using methods from Low Flow Studies (IH, 1980).Corresponding flows for each sorted habitat value were retained as in the following table:
Table 3.2 Output for combined habitat /flow duration curve
• Percentile habitat corresponding
flow
ft 0.50 31.77 28.007
•0.70 0.89
31.75 31.75
27.808 27.808
• 1.08 31.75 27.808
1.47 31.75 27.808
• 2.01 31.7 28.801
2.52 31.7 27.312• 2.94 31.66 27.014
3.52 31.63 26.716• 3.91 31.58 29.794
4.45 31.57 29.894• 5.00 31.52 25.723
5.50 31.47 30.788• 6.00 31.47 25.325
• 6.47 31.46 30.887
7.01 31.42 24.829
•
3.4. Single year duration curvesHabitat duration curves for single years may also be calculated using the IH PHABSIM software, by extracting individual years of data to separate files and passing each individually to the HABDC program. Results for two life stages for brown trout on theWye and Lamboum are presented in Appendix D (figures D6-D12).
•
EA Technical Report WI9 13•
3.5. Flow and habitat spell analysisSoftware to perform flow/habitat spell analysis is not provided with the standard IH orUS releases of PHABSIM. IH has in-house software for low-flows analysis runningunder the Unix operating system on Sun workstations. For this project, individual timeseries of habitat for life stages of brown trout in the Wye and Lambourn were treated as'flow data and transferred in the correct format to the IH Unix system. When data areinput from a file (as opposed to the NRFA or FRIEND river flow archives), only onedataset (i.e. one file) may be analysed per program run. Theprogram is able to interpolatemissing values if required, alternatively it can ignore years where there.are more than acertain number of missing values. For this project the threshold was set to a whole year,to ensure that years were never discarded and interpolation of missing values alwaysoccurred. Thresholds are specified as percentages of the mean 'flows. Although thismethod works well for flows, it may not be the best way to specify habitat thresholds,indeed Capra et aL (1995) specified percentages of optimum habitat. Nevertheless, forthis analysis, percentages were selected as 100, 80, 50, 30 and 10% of mean flow orhabitat as appropriate.
Output is in the form:
Gauge xx MF 23.538 from 1966 to 1981 minimum duration 1 yield (tMF)100.0SPELL FROM TO(days)455. 17 11 1975 13 2 1977226. 28 6 1973 8 2 1974199. 18 8 1978 4 3 1979178. 29 7 1970 22 1 1971143. 19 9 1969 8 2 1970
Gauge xx MF 23.538 from 1966 to 1981 minimum duration 1 yield (tMF) 80.0SPELL FROM TO(days)420. 5 12 1975 27. 1 1977._ .90. 7 10 1973 4 1 197420. 20 11 1978 9 12 1978
The software only outputs the longest period below the threshold in each year, so doesnot compare directly with the method proposed by Capra et al. (1995).
The spell values are plotted as spell (continuous duration under threshold) on the y-axisand cumulative continuous duration (% of total studied duration) on the x-axis. Thetheory behind this, explained in Capra et aL (1995) is that a highly-impacted river willhave a long curve that has a shallow slope, i.e. a high proportion of long periods belowthe threshold. A curve with fewer days below threshold will be shorter, while a lower orsteeper curve of the same length along the x-axis indicates an equivalent number of daysimpacted, but that there are more periods of a shorter duration. Results are presented inAppendix D. -
EA Technical Report W19 14
•••
4. RESULTS•This section refers to statistics presented in Appendix A, Weighted Usable Area vDischarge curves in Appendix B, time series and duration curves in Appendix C andfurther example charts (flows plotted on habitat duration curves, effects ofmissing valuesand flow/habitat-spell analysis) in Appendix D.Habitat model output was taken directly from Project 82.I, which used suitability data developed by IFE Wareham. The only exception was for brown trout on.the.rivers lichen and Lambourn, where the habitat model was re-run using suitability data collected by thethen Wessex NRA, as outlined below.
•4.1. River Exe at Warren FarmThree gauging stations were considered for transfer to the study site: Excat Pixton, Brayat Leehamford and Bark at Brushford. Although it has a larger catchment, the latter wasconsidered to be the only station with a natural regime comparable to the study site, sowas chosen. A complete time series for the 13rushford station from 1982-1992 wasavailable from a previous study (Johnson et al. 1994b), so was used for thisanalysis.
•The habitat model for the Exe was re-run and with the hydraulic model calibrated to halfthe low-flow limit used in the B2.I study, was still within US National BiologicalService(NBS) guidelines of 0.2x the lowest calibration flow. As the time series isa best estimate,rather than direct from a close gauging station, there is perhaps a higher potential error inthe results of the time series analysis for this site. However the methods usedto obtain thebest estimate are current state of the art, and errors in this element are likely to becomparable or lower than those introduced in other parts of the modelling process.
•The most important aspect of this analysis is that the majority of flows are at the lowerend of the calibration, (Figure 4.1 illustrates flow frequencies plotted on the WUA v Qcurve for adult brown trout). For adult and juvenile brown trout (fig C1.4), this results induration curves which are flat to the left of the 10-20% area point, becoming steepertowards the right, and for thc summer curve, flattening slightly around the 95 percentilemark (both summer 95 percentile of 1.6%). The left flat area is caused by flowscorresponding to the peak of the WUA v Q curve, while the slightly higherWUA valuefor adults is a direct consequence of the higher peak of the WUA v Q curve.
•
On the other hand, fry brown trout produce a generally flatter curve with summer habitatsabove the 10% area level being greater than equivalent winter percentiles. This is due tothe model indicating that this life stage has a preference for low discharges.
Spawning trout habitat drops to 0 between 0.4 and 0.5 cumecs, which leads to the rapidtail off of spawning habitat to the right of the 30 percentile level.
4.2. River Wye at Pant Mawr (including comparison with the Exe)Adult and juvenile brown trout show a similar level of optimum habitat tothat on the Exe(20%), but less steep curves (summer 95 percentile of 9.0% area and 7.0% arearespectively). A possible reason for this is that for the Wye, the mean flowis much higherthan the optimum habitat level, while for the Exe it is lower (see Figure 4.1 and B2). The
•
•EA Technical Report WI9 15•
results suggest that abstraction above the Exe site would have more serious consequencesthan abstraction above the Wye.
The 95 percentile fry habitat is also much higher on the Wye, perhaps reflecting theflatter WUA Vs Q curve (i.e. less variable habitat).
Duration curves for invertebrates show some similarity between the NVyeand the Exe, butthe maximum available habitat is higher on the Wye. This is perhaps a reflection of themore simple nature of the WUA v Q curves for these species (compared to trout).However the results are insensitive to the fact that on the Wye, the mean flow is close tothe peak of the curves, while for the Exe, it is well to the left - on the steep portion of thecurve.
The duration curve for Ranunculus again shows a maximum habitat of around 50% oftotal river area, and compared to Leuctridae (A), the central portion of the curve is shiftedto the right.
River Exe at Warren FarmWUA % (total habitat) Numbersof days wrth flow
25 1,000
20 - 800
Adult brown trout WUA
Flow freguendes
15 600
10 4400
5 • ••• o 209
0100 200 300 . . 400-
Discharge (%MF)
WUA % (total habitat)
River Wye at Pant MawrNumbersof days with flow
25
1.600
Adult brown trout WIJA
.•"
-1,400
20
1,200 Flow frequency
15
1,000
800
10-
- 600
400
5-
00000000 200
00000000000000000000000000000000
0
o 100 200 300 400
Discharge (%MF)
Figure 4.1. Rivers Exe and Wye: comparison of flow frequencies and habitat Note: the flowfrequencies are grouped into 25% class intervals, so 12.5 represents flows between 0-25% etc.
EA Technical Report W19 16
•••
4.3. River Hodder at Hodder BankAs with the Exe, the closest gauging station to the site (71002, just below StocksReservoir), is a significant distance (— 15km) from the site. In this case it is above thesite, with several tributaries joining the river system in-between. A record for a moresuitable station, Hodder at Hodder Bridge (71008), is currently being investigated fortransfer to this site, assisted by the methods outlined in III Report 108.
•
4.4. River Blithe at Hamstall Ridware. -
The river has a relatively constant base flow, with some very high but short peakdischarges which greatly exceed the mean flow. The flows are 'un-natural', beingdownstream of Blithefield reservoir, but with the gauging station close to the study site,using the gauged flows will provide an accurate assessment of the conditions experiencedat the study site. Unfortunately, there is a high number of missing flows (591, 10% of thedataset), and (364 flows above the range of the model, a further 6% of the dataset).
•
The model shows high levels of adult and juvenile roach habitat for all percentiles, andhigh levels of roach fry habitat for all but thc most infrequent events. From the WUA v Qcurve for fry, it can be seen that the high model limit corresponds to a lower WUA thanthe low model limit. This indicates that the drop off of habitat above 90% is due to highflows. The spawning habitat duration curve is influenced primarily by the distribution offlows in the spawning months, as the WUA v Q curve increases gradually with flow.
Time series analysis suggests that adult dace habitat is extremely limited, particularly inthe summer months, suggesting that combinations of adequate velocity and depth areunavailable. Juvenile dace tolerate lower depths and velocities than the adults, so thecorresponding curves show rather more habitat at intermediate percentiles, however thedrop off at low flows is still significant. Increasing the model simulation limits in the highflow range would show increasing levels of adult habitat, however the magnitude of thischange is not known. Fry dace are shown to tolerate lower velocities well, and stillexhibit available habitat at the 95 percentile suggesting that higher flows perhaps makemarginal habitats available at the edges.
Invertebrates show generally high levels of habitat, reflecting the high levels of habitatthroughout the modelled range.
•••411
••••
EA Technical Report WI9 17
Adult Dace
50....... I Summer Winter
. I 20
10
5
2
1
0 5WUA
(%
habitat
available)
0 2
0.15 10 20 30 50 70 80 90 95
% of time exceeded
Figure 4.2. River Blithe: adult dace habitat
4.5. River Hellen u/s of HighbridgeFor brown trout, time series analysis was first performed using the suitability datagathered by IFE for R&D Project B2.I. However as can be seen from the WUA vdischarge relationship plot (Figure B4, 'IFE B2.I curves), this results in there beingvirtually zero habitat at mean flow and above. It was suggested that this situation was notrealistic, and in particular, the habitat suitability curve fordepth was suggesting that largeareas of the river were 'unsuitable.
Instead, the habitat model was re-run using brown trout suitability data collected by thethen NRA Wessex Region for applied studies on chalk streams in south-west England(Johnson et aL 1993b, 1994b, 19956). This produced habitat-flow relationships whichwere likely to be more-realistic; these relationships were then-used for subsequent-timeseries analysis.
The 'Wessex' curves suggest that the river provides suitable habitat for adult brown troutat all flows. Absolute levels of fiy/juvenile habitat are similar, but slightly lower.Spawning habitat is high at the optimum spawning flow (15% of area), but suffers aprogressive reduction beyond the 10th percentile, reaching 1.4% area at the 95 percentilelevel. An examination of the WUA v Q relationship and the suitability data suggests thatlower habitat levels are caused by higher flows, and particularly deeper water (see section4.8.2 for discussion).
EA Technical Report WI9 18
WUA (% total habitat)
is
14 •
12
10 I.
SpewOs. Imo ksb4a4
Flow (%MF)
140
ICO
SO
"
40OH1 In 01/11a2 01/11173
Figure 4.3. River Itchen: time series of habitat for spawning brown trout, November and41 December 1970-73.
The invertebrates each have high levels of habitat at all flows (50 percentile > 50% area),as does Ranunculus (50 percentile 30% area). Nasturtium has lower maximum habitat(9%), which decreases to <3% at the 95 percentile, probably due to lower suitability indeep water.
4.6. River Lymington at Balmer LawnFor adult brown trout, optimum habitat (10% arca) occurs for 10% of the time. For therest of the time, habitat drops to a low of 0.42% area at the 95 percentile. Juveniles followa similar pattern, only with an optimum of 4% area. Fry levels seem to bevery high, withoptimum conditions occurring for 75% of the time. Beyond this, the reduction in habitatis caused by high flows. Spawning habitat is optimal (5% area) for 10% of the time, anddrops to a 95 percentile level of 0.2% area.
For invertebrates, Gammaridae (A) summer habitat levels are above 10%area for 95% ofthe time. On the other hand, Leuctridae suffer a significant fall off on habitat,dropping toless than 1% area at the 95 percentile. A glance at the WUA v Q curve suggests that thisdrop is due to low habitat levels at low flows. Nasturtium and Ranunculus show similarduration curves, despite having very different WUA v Q relationships (see section 4.10for discussion).
4.7. Millstream at I.F.E East StokeAdult dace show a marked decline in habitat from a high optimum (40%) above the 5thpercentile in summer and 30th percentile in winter, with habitat dropping below 1% oftotal habitat at 75% (summer) and 92% (winter). As on the river Blithe, this is probablyduc to the adult's preference for water above 40 cm depth and 30 cm/s velocity.Juvenilesshow a similar tail-off, but from a lower optimum (15% of total habitat). The summer andwinter curves are more similar than for adults. For fry, as with dace and other species onother rivers, summer habitat is maintained at a more consistent level, even at low flows.Spawning habitat is consistently high, except at low flows, reflecting its wide preferencecurves.
Adult, juvenile and fry roach again show relatively high levels of habitat at low flows, and greater summer habitat at medium percentiles. This reflects their tolerance of lower flows, specifically lower velocity. Lower fry habitat at the 95 percentile may represent
EA Technical Report W19 19
lower tolerance of high flows. Low base levels of spawning habitat may be influenced bythe narrow substrate preference curves, along with water shallower than 50cm. As levelsof vegetation are not specifically modelled, results should be treated with caution.
The invertebrates show very shallow curves, demonstrating relatively insensitiveresponse to flow, except at very low flows. High levels of Ranunculus habitat reflect thedistribution of the WUA v Q curve optimum around the meanflow.
4.8. River Lambourn at Hunt's GreenAs with the Itchen, the 'Wessex NRA' suitability data forbrown trout in chalk streamswas used in the time series analysis. Compared to output using the IFE B2.I curves, useof these curves indicates greater habitat availability at high flows, and slightly less habitatat the lowest modelled flows.
The high base flow on the Lambourn produces 95 percentile habitat levels that are arelatively high proportion of the optimum, for all brown trout life stages. Similar resultsare obtained for invertebrates and macrophytes.
Adult Trout
I
Summer Winter
50
30
20
10
5
3
2WUA
(%
habitat
available)
5 10 20 30 50 70 80 90 95% of time exceeded
Figure 4.4. River Lambourn: habitat duration curve for adult brown trout
Comparison of Lambourn and lichenA combination of a high base flow, coupled with the natureof thc 'Wessex NRN browntrout suitability curves, leads to quite flat habitat duration curves for adult, andfry/juvenile life stages. In contrast, spawning habitat at high percentiles (e.g. 95) is stillhigh on the Lambourn, but declines on the Itchen. Possible reasons for this include eitherthat it is not appropriate to apply the 'Wessex' brown troutcurves to rivers the size of theItchen, or that there was indeed little suitable spawning areaat the study site.
4.9. River Gwash at BelmesthorpeFor both adult and juvenile brown trout, optimum habitat levels are present for 30% ofthe time in summer and 20% in winter. As with other rivers, the magnitude of optimumhabitat area for adults is greater than for juveniles. Examining the WUA v Q curve for fry
EA Technical Report W19 20
•
•
•trout shows a dramatic decrease in habitat area with increasing flow, with an optimum of 10% of total area at 25% MF, declining to 1% area at 100% MF. This is reflected in the duration curve, which shows a 95% summer percentile of just 0.2% area available. Dropoff for spawning habitat is even more dramatic, with a dramatic and rapid reduction inhabitat available above the 50th percentile.
Invertebrate and macrophyte habitat levels are maintained well, and at the 95th percentileshow a relatively small drop from the optimum.
4.10. River Lymington: habitat duration curves combined with flowdata(See figures Dl-D3)
41' The duration curves for Ranunculus, Nasturitum and brown trout fry on the Lymingtonillustrate the care that must be taken in using duration curves to interpret habitat timeseries.Plotted on the duration curve is not only habitat, but the flows corresponding to eachhabitat percentile (note that this is not thc same as flow exceedance percentiles). Therelationship for Nasturtium is relatively simple, with high flows producing low habitatand vice-versa. A similarly shaped curve is produced for Ranunculus, yet in this case,high habitat levels come from intermediate flows, intermediate habitat levels from acombination of high and low flows, and the lowest habitat from low flows. Likewise thebrown trout fry curve produces similar high habitat from intermediate flows, but in thiscase the lowest habitat is caused by high flows.
4.11. River Gwash: plotting missing values on time series and durationcurves(See figures D4&5).The IH duration curve suite of software considers habitat values arising fromflows out ofrange of the model (i.e. at the low and high flow ends) to be 'undefined. The significanceof periodic very low and high flow events for PHABSIM analysis has never beenthoroughly analysed. There has been some suggestion (Milhous 1990b), that the biotawould not respond to events shorter than a month, but with no clear supporting evidence;this was then used to justify the choice on a monthly time period for time series analysis.Alternatively, one could suggest that acute low flow periods of only a few days could alsohave an effect on a stream population. Clearly the use of monthly mean flows, ifappropriate would be a way of drastically reducing the number of out of range flows.What is certain is that ignoring out-of-range flows is not a very satisfactory solution.Other possible solutions include:
• Ensuring a widest possible range of calibration flows, resulting in fewestpossible out-of-range flows.Ensuring the most accurate possible calibration, from accurate field measurementsand care in the calibration itself. Aside from giving better results over the range ofcalibration flows, a good calibration will be more forgiving when extrapolated outsidethe calibration flows.
• In the absence of the above, the potential for simply setting out-of range flows to thelimits of the model is examined here.
•
EA Technical Report WI9 21•
The rationale behind setting out-of range flows to model limits is simply that whenconstructing a duration curve of habitat, inclusion of habitat at high exceedancepercentiles (set to model limit) will result in a more accurateduration curve for the rangeof flows within the model. For example on the river Gwash, there are 132 summer flowsand 385 winter flows outside model range, a total of 8.8%of the data. Considering adultbrown trout, as habitat at high flows is at a low percentage of the optimum, ignoring thisdata must have an impact. The impact on the duration curve will be in two forms,incorrect percentiles at the extreme right of the duration curve and inaccurate percentilesfor ALL OTHER HABITAT VALUES.
As the WUA at the high limit of the model is lower than the WUA at the lower limit, onecan be sure that all these flows will correspond to the right hand limit of the durationcurve chart. Figure 135illustrates duration curves for adultbrown trout with and withoutthe out-of range flows set to the model limit.
From Table 2.1, it can be seen that 132/5844 = 2% of summer flows and 7% of winterflows are beyond the high limit of the Gwash habitat model. Figure 135.2illustrates theeffect of their inclusion, set to the habitat value at the highlimit. The flat portions on theright of the summer and winter curves reflects these valueswhich are estimates, but stillincorrect. However due to the inclusion of all flows, valuesto the left of the 98 percentilefor the summer graph, and 93 percentile for the winter graph, are at their correctpercentiles. This is not true for the duration curve in fig 1)5.1.
Comparing the two graphs, one can now see for example that when the missing data isincluded, 80 percentile summer habitat is reduced from 4%area to 2% area.
River Gwash, adult brown troutWIJA % (total habitat)
20
15
Adult
10 -
High model limit
gives lower habitatvatue than lowermodel limit
50 100 150 200
Discharge (%MF)
Figure 4.5. River Gwash: WUA v Q curvefor adult brown trout
It should be noted that this process as outlined above wasonly valid when the habitatvalue of the limit beyond which the out-of range flows are incorporated is lower than thehabitat value for the other limit of the graph. Otherwise, a series of 'incorrect' values willbe incorporated into some central portion of the duration curve, which is of little use.
5-
EA Technical Report WI9 22
••
To conclude, to avoid bias in the construction of duration curves, it is necessary to ensurethe widest possible model calibration. In the absence of the widest possible calibrationlimits, a well-calibrated model could be extended beyond the limits currently advised inorder to minimise bias in the duration curve.•4.12. Lambourn and Wye: comparison of single - year duration curves(Sec figures D6-DI I)In order to try to overcome the loss of temporal precision that comes from_producingaduration curve from a long time series, yearly duration curves of flow, adultand juvenilebrown trout habitat were plotted for the Lambourn (lowland site, high base flow) andWye (upland site, flashy catchment). The time period was 1966-1977. The Wye shows aremarkably stable yearly habitat regime, while the habitat on the Lamboum seems to beclearly impacted during 1976, and to a lesser extent 1977. In both cases, 95 percentilehabitat shows less variation than 95 percentile flow.
Table 4.1. Lambournand Wye: Variationin 95percentileflow andhabitat•
Wyeflow (% MF) adult (WUA)fry (WUA)
Lambournflow (% MF)adult (WUA)fry / juv(WUA)
min.
max.4.8 0976) 23 (1967)9.0 (1974) 10 (1963)6.7 (1974) 9.4 (1976)
22 0976) 80 0968)14 (1976) 24 (1968)10 (1976) 23 0968)
4.13. Lambourn and Wye: flow and habitat spell analysis(See figures D12&D13)The graphs show that the narrower range of habitat data (compared to flowdata) meansthat it is especially important to choose the correct thresholds. Thresholds within 10% ofeach other may give widely differing results.
The Lambourn is dominated by a few years in which habitat drops below the thresholdsfor long periods. For adult trout, the curve for the 100% mean habitat threshold is high(up to 550 days) and long (up to nearly 50% of total duration). The fry/juvenile troutcurve is shorter and lower, suggesting that the adult trout may be most impacted by lowflows.
•On the Wye however, for adult trout the cumulative duration of flows below 100% meanhabitat is shorter (8% of time period) and the longest time period is also shorter(75 days).
•Comparing flow and habitat deficit periods, it appears that habitat deficits seem to beshorter but more frequent.
• Modification of the IH flow-spell software to considering all periods below threshold ineach separate year, plus the ability to specify percentages of optimum habitatwould forma useful part of future PHABSIM development work.
•
•
•EA Technical Report W19 23
•
5. CONCLUSIONS AND RECOMMENDATIONS
Analysis of time series is a major step towards an understanding of how physical habitatmay limit or alter river wildlife, as the biota will be influenced not only by currentphysical habitat, but also past physical habitat events. Unlike the wealth of literature onprocedures for other parts of the PHABSIM process (e.g.collection of suitability data,hydraulic modelling), information on time-series procedures is limited. This reportprovides a methodology for application of physical habitat time series in UK Fivers.
The additional programs WUAQT and HABDC written at IH and provided on the EArelease of the PHABSIM software provide the basis for time series analysis of habitatdata. WUAQT produces a time series of habitat from a time series of flow, whileHABDC produces duration curves, of habitat for a species / life stage, flow and totalhabitat.
This initial study has concentrated on the temporal relationships between daily physicalhabitat and daily mean flow for nine quite different UK rivers. In order to aid comparisonbetween rivers, habitat has always been expressed as a percentage of total habitat area inthe river in question, and flow as percentage of the mean flow. Due to the nature of thedata collection program, only broad comparisons between river types was possible.Comparisons between sites has been restricted by the different species present, and in thecase of brown trout, the lack of UK-specific suitability curvestransferable to all sites.
To some extent, a river's habitat time series will reflect the flow regime. For example inthis report, the flashier upland streams show steeper habitatduration curves than do chalkstreams. However, habitat for a life-stage is also fundamentally linked to the shape of theWUA v Q curve. Depending on the shape of the curve (combined with characteristics ofthe flow regime), on some rivers, small changes in flow may produce large changes inavailable habitat, in contrast on other rivers, large changes in flow may produce smallchanges in habitat. Figure 5.1 illustrates the properties of a WUA v Q curve that mayparticularly influence habitat time series. A fundamental property of PHABSIM is thatthere will be a flow corresponding to an 'optimum' habitat, this leads to significantdifferences in the shape of a habitat time series compared to the flow time scries for ariver.
Habitat duration curves are a valuable analysis tool, particularly in the assessment ofalternative flow regimes. However their limitations, particularly thc influence of missing /out of range data, and the potential relationship between one habitat value and two ormore flows must be recognised. When applying PHABSIM,all possible attempts shouldbe made to ensure the widest and best model calibration tominimise the number of out-ofrange flows.
Depending on the combination of flow regime and WUA v Q relationship, seasonalhabitat duration curves may be similar to or different fromwhole-year habitat curves.
In this study, habitat was highest for the invertebrate groups, with optima reaching over 50% area for Leuctridae and Polycentropus flavomaculatus on the river Wye, and for Ephemeridae and Gammarus pulex on the river Itchen. Although invertebrate habitat in
EA Teclmical Report W19 24
this study was generally high, even at high (95+) exceedance percentiles, this is partly areflection of the broadly-based nature of the suitability curves. It would not preclude thepossibility of greater habitat reduction for other specific invertebrate species in the studyrivers. Habitat for macrophyte species was also consistently high.
•Habitat calculated for fish species was much more variable, with optima varying between40% area (adult dace in Millstream) and 1.6% area (spawning trout on Gwash).Variability in habitat, calculated as (5 percentile habitat)/(95 percentile habitat) variedbetween 1.3 (adult trout on Lamboum) to 1300 (adult dace on Blithe and Millstream). On
111 the cyprinid streams, dace habitat showed a consistent, severe reduction in habitat at low flows, while roach habitat levels were maintained at low flows. For brown trout, dace androach, spawning habitat levels were difficult to interpret, and widely differing WUA v Qrelationships produced comparable spawning habitat duration curves. This suggests thatsignificant information was being lost in the construction of the duration curves.
Differences between life stages on the same river showed some consistency, often whereadult habitat was highest, juvenile habitat was lower and fry habitat lowest, and vice-versa, although there were some exceptions. For brown trout on the Exe, Wye,Lymington, roach on the Blithe and Millstream and dace on the Blithe, fry habitat levelsat high exceedance percentiles were greater than corresponding percentiles for adults,consistent with a greater preference for fry for low flows. Fry dace on the Millstream andbrown trout on the Gwash showed the opposite tendency.
Differences in the known territory requirements of different life stages would provide auseful comparison, especially regarding the areas required for spawning.
This report has highlighted the requirement for greater information on transfer of habitatsuitability curves between different rivers (both within and between ecological groups).
•
Recommendations for software enhancementCertain additional features in the HABDC software would be highly desirable, inparticular the ability to produce separate duration curves for each year in a time series. Itwill also be possible to enhance the IH flow spell analysis software to produce data on allperiods below threshold in each year of a time series. This could be combined with thetransfer of the software to the PC platform, in preparation for its incorporation into afuture UK version of PHABSIM.
Recommendations for further data analysis
More information is required on how habitat varies in rivers of the same ecologicalgroup, and to compare rivers of different ecological groups. This could be combined withan investigation of 'regional' suitability curves.
•Another useful line of work would be a sensitivity analysis of the effects of alteringpreference curves on both WUA v Q relationships in different rivers, and time seriesmethods such as duration curves. One way of achieving this could be with a set of'parameterised' suitability curves, specified by mean and standard deviation.
Sensitivity analysis of the influence of systematic errors in the flow time-series on habitatduration curves (and low habitat spells) would also be useful, especially given the•EA Technical Report W19 25
requirement to sometimes transfer a flow time-series from a gauged to ungauged site.Topics in forthcoming 11-1research on low flow estimation such as flow time seriessynthesis and seasonal duration curve estimation are highly relevant to this work (NRAR&D Note 330, Gustard et al. 1995).
Better ways must be sought to display information such as habitat duration and spellcurves. There is certainly scope for further modification and adaptation of existing lowflow' analysis methods as applied to quantitative habitat analysis.
Although not attempted in this report, there may be some merit in applying regionalhabitat suitability data to a river model even where species are not found (or rarelyfound), for example brown trout in the Millstream. This would indicate whether physicalhabitat, or some other factor such as water quality or inter-species competition waslimiting.
This report has only touched on the influence of critical periods of habitat. The use of'effective habitat' analysis (Bovee 1982) should be investigated in the UK as a precursorto linking WUA more closely with biomass estimates fromfish surveys.
EA Technical Report W19 26
An example WUA v Q curve
•30 30
25 25
20 20
• 15 15
•10• 10
5- 5
-
Absolute values of habitatwuA i4 (ln babrtat) WUA%(totaltablet)
-
• oo so 100 150 200
•
so 100 150Chscharge(%MF) Oncharge(SWF)
Shape of cone at low flows Shape of curve at highflows
WUA%(totaltabnat) WSIA%(totalballet)
200
•30
25
30
25
• 20
20
• 15•
15
10
10
•
5 -
5 -
0 0
'
0
o so leo 150 200 0 50
t
100 150 200Discharge(%MF) Discharge(%MF)
Slope of curve Doubte-peaked areW1JA%(totalbat>nat) WUA%(totalbabas!)30 30
•25 25
20 • 20
15 - 15
10 _Apr 10 45 - s -
o ' oo
100 150 200 o 100 150 200
Discharge(%MF) Ducharge(%W)
Figure 5.1. Variations in WUA v Q relationship
*EA Technical Report WI9 27
REFERENCES
Armitage, P.D. & Ladle, M. (1989) Habitat preferences of target species for applicationin PHABSIM testing. A report to the Institute of Hydrology,pp 56.
Bartholow, J.M. & Waddle, TJ. (1986) Introduction to Stream Network Analysis.Instream Flow Information Paper No. 22. U.S.D.I. Fish and Wildlife Service Biol. Rep.86(8).
Bartholow a al. (1993). A Salmonid Population Model with Emphasis on HabitatLimitations. Rivers 4, 265-279.
Bovee, K.D. (1982) A Guide to Stream Habitat Analysis Using the Instream FlowIncremental Methodology. U.S. Fish and Wildlife Service, lnstream Flow InformationPaper No. 12, FWS/OBS - 82/26.
Boyce, K.D. (1986). Development and Application of HabitatSuitability Criteria for Usein The Instream Flow Incremental Methodology, US Fish& Wildlife Service InstreamFlow Information Paper No. 21.
Bovee, K.D. (editor) (1995). A comprehensive overview of the Instream FlowIncremental Methodology. National Biological Service, FortCollins, Colorado.
Bullock, A., Gustard, A. and Grainger, E.S. (1991) Instream Flow Requirements ofAquatic Ecology in Two British Rivers. 111Report No. 115.
Capra H., Breil, H. and Souchon, Y. (1995). A new tool to interpret magnitude andduration of fish habitat variations. Regulated Rivers: Research and Management, Vol 10,281-289.
Cheslak, E.F. and Jacobson, A.S (1990). Integrating the Instream Flow IncrementalMethodology with a Population Response Model. Rivers I 264-288.
Elliott, C.R.N., Dunbar, M.J. and Gowing, I.M. (I 995a). Assessment of the Potential useof the Physical Habitat Simulation System (PHABSIM) to Assess the Impact of FlowModification on Juvenile Fish Habitat in the River Thames. Draft Final Report. Instituteof Hydrology Report to EA Thames Region.
Elliott, C.R.N. and Johnson, I.W. (1995b). Assessment and Design of HabitatImprovement / Restoration Procedures. Institute of Hydrology Report to MAFF, ProjectFD0505.
Gordon, N.D. McMahon, T.A., Finlayson, B.L. (1992). Stream Hydrology, anIntroduction for Ecologists. John Wiley and Sons, Chichester,UK.
Gore, J.A. (1989) Models for predicting benthic macroinvertebrate habitat suitabilityunder regulated flows. In 'Alternatives in Regulated RiverManagement' (eds. Gore, J.A.& Petts, G.E.), CRC Press Inc, Baton Rouge, Florida, 253-265.
EA Technical Report W19 28
Gore, J.A. & Nestler, J.M. (1988) Instream Flow Studies in Perspective. RegulatedRivers: Research & Management 2, 93-101.
Gustard, A., Cole, G., Marshall, D.C.W., & Bayliss, A.C. (1987) A study ofcompensation flows in the United Kingdom, Institute of Hydrology Report No 99.
Gustard, A., Bullock, A, Dixon, J.M. (1992). Low Flow Estimation in the UnitedKingdom. Institute of Hydrology Report No. 108.
Gustard, A., Young, A., Bullock, A. (1995). Scoping Study to Develop Design Tools forLow Flow Estimation. NRA R&D Note 330. Institute of Hydrology.
Harvey et al. (1993). A Physical Process-biological Response Model for SpawningHabitat Formation for the Endangered Colorado Squawfish. Rivers 4, 114-131.
Hearne, J.W., Armitage, P.D. (1993) Implications of thc annual macrophyte growthcycle on habitat in rivers, Regulated Rivers, in press.
•
Heggenes, J., (1990) Habitat Utilization and Preferences in Juvenile Atlantic Salmon(Salmo solar) in Streams. Regulated Rivers: Research and Management, Vol.5, 341-354.
Institute of Hydrology, (1980) Low Flow Studies. Institute of Hydrology Report.•
Jager et al. (1993). An Individual-based Model for Smallmouth Bass Reproduction andYoung-of-year Dynamics in Streams. Rivers 4, 91-113.
IDJohnson, I.W., Elliott, C.R.N., Gustard, A., Armitage, P.D., Beaumont, W., Ladle, M.,(1991) Data requirements and data collection procedures for application of the instreamflow incremental methodology in the UK, Institute of Hydrology report to the NationalRivers Authority.
•Johnson, I.W., Elliott, C.R.N., Gustard, A., Armitage, P.D., Ladle, M, Dawson, F.H.,Beaumont, W. (1993a) National Rivers Authority Project 82.I Ecologically acceptableflows Report. Institute of Hydrology report to the National Rivers Authority.
Johnson, I.W., Elliott, C.R.N., Gustard, A., Clausen, B. (1993b) River Allen instreamflow requirements, Draft final report, Institute of Hydrology report to the National RiversAuthority.
Johnson, I.W., Elliott, C.R.N., Gustard, A., Yaqoob, S., Fairhurst, A., Blackburn, J.H.,Armitage, P.D., and Ladle, M. (1993c). Quantitative Environmental Assessment of RiverFlood Defence Schemes. Final Report. MAFF Project FD0503. Institute of Hydrology.
•
Johnson 1.W., Elliott C.R.N. and Gustard A. (1994a). A guide to field data collection techniques for application of the Instream Flow Incremental Methodology using thePhysical Habitat Simulation System (PHABSIM). Institute of Hydrology.
•EA Technical Report W19 29
Johnson I.W., Elliott C.R.N., Gustard, A. (1994b). Ecologically acceptable flow regimes:River Bray at Leehamford, River Bade at Perry Weir. Final Report to National RiversAuthority South Western Region.
Johnson I.W. and Law, F.M. (I 995a). Computer models for Quantifying the Hydro-Ecology of British Rivers. J.CIWEM 9, 290-297.
Johnson I.W. and Elliott C.R.N. (1995b). Habitat Modelling and Instream FlowRequirements: River Piddle. Draft Project Report. Instituteof Hydrology:
Johnson I.W., Elliott, C.R.N. and Gustard A.G. (1995c). Modelling the effect ofgroundwater abstraction on salmonid habitat availability in the River Allen, Dorset,England. Regulated Rivers, Research and Management 10229-238.
Jowett G. (1992). Models of the abundance of large brown trout in New Zealand rivers.North Am. J. Fish. Manage. 12, 417-432
Lamb, B.L. & Doerksen, H.R. (1987). Instream Water Use in the United States-WaterLaws & Methods for Determining Flow Requirements. National Water Summary 1987-Water Supply & Use: lnstream Water Use, 109-116
Mann, R.H.K, (1973) Observations on the age, growth, reproduction and food of theroach Rutilus rutilus (L.) in two rivers in southern England.J. Fish Biol. 5, 707-736
Mann, R.H.K, (1974) Observations on the age, growth, reproduction and food of thedace Leuciscus leuciscus (L.) in two rivers in southern England. J.Fish Biol. 6, 237-253
Mathur et al. (1985). A critique of the instream flow incremental methodology.
Milhous, R.T. (1986). Development of a Habitat Time Series.Journal of Water ResourcesPlanning and Management, 112, 145-8
Milhous, R.T. (1988). The Physical Habitat-Aquatic Population Relationship.Presentation at the 1988 Meeting of the Colorado-Wyoming Section of the AmericanFisheries Society. Unpublished.
Milhous, R.T. (1990a). User's Guide to the Physical HabitatSimulation System - Version2. Instream Flow Information Paper No. 32. U.S.D.I. Fishand Wildlife Service Biol. Rep.90
Milhous, R.T., Bartholow J.M., Updike M.A. and Moos, A.M. (1990b). ReferenceManual for Generation and Analysis of Habitat Time Series - Version II. Instream FlowInformation Paper no. 27.S Fish and Wildlife Service Biol.Rep. 90(16).
Mountford, 0., Gomes, N., (1990) Habitat preferences of the River Water Crowfoot forapplication in PHABSIM testing. ITE report to the Instituteof Hydrology.
National Rivers Authority (1993). Low Flows and Water Resources. NRA, Bristol.
EA Technical Report WI9 30
Nestler, J.M., Milhous, R.T., Layzer, J.B. (1989). Instream Habitat Modeling Techniques.In: Alternatives in Regulative River Management, J.A. Gore and G.E. Petts (Editors).CRC Press, Boca Raton Florida, pp295-315.
Newman, A.T., Symonds, R.D., (1991) The River Allen - a case study, in Procs. BHS 3rdNational Hydrology Symposium, Southampton.
•
Orth D.J. (1987). Ecological Considerations in the development and application ofinstream flow habitat models. Regulated Rivers: Research and Management1, 171-181
Pens, G.E. (1989). Methods for Assessing Minimum Ecological Flows in British Rivers.BHS National Hydrology Meeting- Prescribed Flows, Abstraction & EnvironmentalProtection.
Petts, G.E. (1990). Application of the Instream Flow Incremental Methodology in theUnited Kingdom. Report by the Freshwater Environments Group, Dcpartment ofGeography, Loughborough University of Technology.
Scott, D., Shirvell, C.S. A Critique of the Instream Flow Incremental Methodology andObservations on Flow Determination on New Zealand. Department of Fisheries andOceans.
•
Souchon, Y., Trocherie, F., Fragnoud, E., Lacombe, C,. (1989) Instreann FlowIncremental Methodology: Applicability and New Developments. Revue Des SciencesDe L'Eau, 2, 807-830.
Trihey, E.W. & Wegner, D.L. (1981) Field Data Collection Procedures for use with thePhysical Habitat Simulation System of the Instream Flow Group. Cooperative InstreamFlow Service Group, Fort Collins, Colorado.
•Vogel, R.M., Fennessey, N.M. (1995). Flow duration curves II. A review of applicationsin water resources planning. Water Resources Bulletin 31, 1029-1039.
•Winfield, I.J., Towsend, C.R., (1991) The role of Cyprinid fishes in ecosystems.Cyprinid fishes, systematics, biology and exploitation, Ed Winfield, I.J., Nelson, J.S. Pub.London, Chapman & Hall, pg 552-571.
•
Young, A.R. (1992). Micro Low Flows 1.31. User Guide and Technical Reference.Institute of Hydrology.
EA Technical Report WI9 31•
APPENDIX A. SUMMARY OF GAUGED FLOW DATA ANDHABITAT STATISTICS
Exe(MF0.183,X=flow)Trout (X=WUA)AdultJuvenileFrySpawning (Nov-Dec)Wye (MF=1.7 cumces,X=flow)Trout (X=WUA)AdultJuvenileFrySpawning (Nov-Dec)Blithe(MF1.04 cumecs, X=flow)Dace (X=WUA)
AdultJuvenileFry
Spawning(March-April)Roach (X=WUA)AdultJuvenileFry
Spawning(April-June)
Itchen(ME5.40 cumecs, X=Q%MF)Trout (X=WUA)
AdultFry / Juvenile
Spawning (Nov-Dec)Lymington (ME 1.01comas. X=4::)%MF)Trout (X-TWUA)
AdultJuvenileFry
Spawning (Nov-Dec)Millstream (MF 0.93cumecs, X=Q%MF)Dace (X=WUA)AdultJuvenileFry
Spawning (March-April)Roach (X=WUA)AdultJuvenileFry
Spawning(April-June)
SummerX mean
10.79.6115.9
67.1
15.3 13.5 14.0
66.8
2.274.168.7114.2
19.018.023.0069
90 _5
-23.617.1
53.9
2.77 1.33 26.5
14.37.065.51412
17.016.516.0
0.635
X5
22.918.919.7
20.0 19.4 18.1
12.59.7211.119.0
23.922.424.31.92
-25.9 21.5
10.2 3.85 31.5
40.113.38.2
50.2
20.620.923.21.08
XSO
31.7
7.016.8818.7
33.9
15.8 12.2 15.0
48.0
0.353.969.4512.0
19.318.024.00.48
84
25.0 19.3
11.2
0.890.89
29.76
59.3
7.587.145.7647.6
18.418.119.00.57
X95
8 2
1.60 1.61 8.60
6.91
8.96 7.04 7.37
30.0
0.010.674.361.74
13.914.018.00 17
53.0
17.78.21
4. 77
0.260.26
7.6
10 2
0.030.452.8218 3
7.397.444.920.44
X51X95
1412
2.3
2.2 2.8 2.5
130015
2.5106
1.7 1.614 7 1
1.5 2.6
4015
4.1
130030
3.02.7
2.8 2.8 4.72 5
Winter X mean
15.3 12.9
2.2
133
14.9 12.0
2.06133
6.33 5.83
20.7 19.5
HO
22.815.57 98146
5 44 2 03
1.66
21.9 6.13
15.1 14.3
X5
.-n/a
19.2
20.0 19.4
4.04
30.3 11.2
25.3 24.5
25.9 21.5 14.2
10.54.1
5.17
40.413.0
20.4 20.7
XSO
100
_16.2 12.3
1.0773.1
14.3 10.9
2.06 67.4
1.60 5.77
21.0 19.5
105
24.316.97.3883.8
5.51.73
0.82114
32.0 7.45
17.2 15.3
22.6
33.0
X95
27.9
4.34 3.90
0.08
10.05.84
0.04
0.031.12
14.9 14.8
59.0
16.47.131.3914.3
0.42 0.42
0.13 25.3
0.45 0.44
6.97 7.09
X51X95
n/a 4.9
..)
3.3
100
100010
1.7 1.7
1.63.010
25 10
22
90 30
2.9 2.9
444444444444•4
4
•4
•4
•4
4
••••
EA Technical Report W19 32
DLa m bourn(N4 F=1.31cumces, X=Q•41‘111Trout (X=WUA)
105.3
99.4 34.1
94.9
76.1 40.2
Adult 24.5 26.5 25.5 19.8 1.3 22.6 26.4 23.2 17.1 1.5D Fry / Juvenile 23.1 26.9 24.3 15.9 1.7 20.2 26.6 20.4 12.2 2.2
Spawning (Nov-Dec)
12.1 14.2 12.0 9.99 1.4
D Cwash(MF=0.90tu men, X=Q%MF)
85.1
69.8 30.7
115.5
85 31.9
• Trout (X=WUA)
Adult 10.4 17.7 9.33 3.50 5.0 9.06 17.5 7.71 1.51 12• Juvenile 6.5 9.09 6.59 2.33 3.9 5.77 9.07 ...5.73 1.28 7.1
Fry 5.25 10.8 4.86 0.16 68
D Spawning (Nov-Dec)
1.60 2.61 1.98 0 nia
•••••
EA Technical Report WI9 33
APPENDIX B. RELATIONSHIPS BETWEEN WEIGHTEDUSABLE AREA (WUA) AND DISCHARGE (Q)
-
34
• WJA(11510X4o)
WUA (m2/1000m)
Brown TrOUt(IFE B2A Curves)
WUA (% total available habitat)
Brown Trout (WE Bit Cuntes)Wua % (WM50480
1.CCO
25
AMA AMA
•
800
Arena 20
Juvenis
• 503
15
SP4444800 -•
SpaTenng• 400
0 10•
•200 0
5
L.
•
0
0
0 05
15
2 0200
400600 800•
DuscharGA(0454040
Nunn, (%MF)
•
Invertebrates
Invertebrates
• WUA(AW103041)
WUA(%lotshab01)
1KO
50
CNorop Ch 10/0P
• 1.403 1.200
Mut& 40
I.e uar
•
•• 1OW
30
-
20
• &CO
403
10
• 200
0
•
000 5
Dwol004(44.0.04)15
20200
400500orscrwoo(/8519
eoa
•
Total habitat
WUA080000TO
• 4=0
TotMb
•
3020
• LOCO
• LOCO-
00 05 15 2
Chschamo(54A455)
Figure Bl. River Exe: habitat vs discharge relationships
EA Technical Report W19 35
WLIA (m./1000n)
3.0D3
WUA (m2/1000m)
Brown Trout (WE 82.1 Curves)
WUA
25
WUA (% total available habitat)
Brown Trout (IFE 82.1 Curves)(Io!al 444431)
AduilAdv.
2.500
Juvenal, 20
Intonate
2003r
• Fry
15 -
1.500
SP ‘4:4'4410
Spating
10
1.000
4-
500
5 -
o
0
o 1 13 • 5 50 100150 203 250 300
Dtscharge Ions)
DI$4.41434(14MF)
MIA (rn510004)
12.000
10.000
13.000
6.030 -
2.000
00 I 2 3
13.schato• (area)
Lew
Poty F1av
• Flamm
5
Invertebrates and MacrophyteWUA total +.1,0”
aoLa..=70 -
Poly Flay
• Rarunc
50 1W 150 200 250 303Drscharge(10.1F)
Invertebrates and Macrophyte
eTh
60 -
50
40 -
30
20•
10
0
Total habltatWUA (mU1000m)
16.000
14.000
12.000
10000
8.003
6.003
2.000
00 2 3 • 5
Dacharge (cornea)
Figure B2. River Wye: habitat vs discharge relationships .
EA Technical Report WI9 36
••••••••tt•••••••••••••••••••••••
WUA (m2/1000m)
RoachWUAW/103.3”1
3.500
3.000
2.500
2.00
1500
1,C00
500
000 51I 322.53
own ripe (cuss)
Dacewto, toT1:1003n4
5.000
4,00
3000
2003
1.003
00 SII S22 5
°whole lesassa)
Invaebbntes and MacrophyteWILAW/112C0m)
5000
&OM
1.030
oo0511 522 53
Der." rip (earoks)
Total habitatVIVA (MIID:°m)
WUA (% total habitat)
RoachMIA1%MS N•3120
320
300
300
Mut
inia
•SpoWnec•
33
25
20
15
MIA
Amnia
07
SCOActo
10
5
050100193203250
Osthrge (%1.49
Dac•WUA (% mad harm)
Mon
June.%
Spavesre
40
30
10
AtkOl
SSTS@
PT
SPSna0
10
30050100150203250
Csulurp (%149
Invertebrates and MeciophyteWUA (%10141liabtag
Gwent/
Poly Flay
5o
40
30
Gower
Poky Flaw
Nt
20
10
ooso100150200250
Dischato• (10.10
River Blithe
14.030
12,030Tot HS
10.030
8.00
LOX
I.033
2.00
0o0 SIIS22 53
0 Ww9t 13200,3s)
Figure 83. Habitat vs Discharge Relationships:
EA Technical Report W19 37
WUA (mV1000m) WUA (% total habitat)
MIA (W/1000m)
5.000
4.030
Brown Trout (Wessex NRA Curves)
%WA (t tots lublet)
30
25
Brown Trout (Wessex NRA Curves)
tha
FrOw
•
ADA
ry/Ju•
3.000
30•041830 20
0404.3433
•
2 000 •
15
IC
0
0
D4020 40 6040100 120 (40
Ordwo• loath)
Man, os kw)
Brown TroutE 132.1 Cries)
Brown Trout ((FE 621 Cries)
WUA 01W1000m)
SALM(% 100 ho.b1011
3.500
20
0001Ada
3.000
2 SCO
4u4•4141
Ns*
2 COO
10
1 500
SOO
0 2 4DO2040 60 SO icein 140 160
Cledwg• (OarneC3)
CtSCIIV40 fth 44)
Invertebrates
Invert abet es
MIA we/1000ov
MIA Esttete nattist)
12.000
70
EpsomE4Sunt
10 000
C.34440/60
carewt.
• SO
40
6.000
30
4 000
20 .
2 000
10•
° 0 2 40020
Dectthaos (agree)40 40 so100120
causens (% 34.)
HO tee
Macrophyles
Macrophytes
WuA wW1000/16
YAM (NoMY hob, 0
7 020
40
RthourcRVILCIL
6.000
Nifty
none5 000
030
0000
20
3 000
2.000
10
1.000
00 2 4 a00 20 a 4010100(20 140 160
Dectwo• (omen
01schoor 111411)
Total habitat
%VW (l1/10:0e)
Total nab
•Dadante lona)
Figure 84. River ltchen: habitat vs DisCharge Relatibiz§hips
EA Technical Report W19 38
20.000
15 000
102:03
3 OM
WUA (m2/1000m) WUA (% total available habitat)
503- r
0o 0 5 1 I 5 2 2.5 3
Disch440. (comma)
Invertebrates
2.500
2.000
1.500
1....
W.th (nWICOOrn)
3.503
3 OCO
•
2.003
1.503
1.000 -
SCO
VILA (% Iota hamm)
30
25
20
• 00 5 1 1.5 2 2.5
Orschwas (cameo)
050 100 150 203 250 300
13Khan (%&IF)
Roust.
Namur
•
40
30
20
Ramon
Nastur.
0
Macrophytes MacrophytesWJA (% total habtra)
50
0o 05 I 1.5 2 2 5 3
Dn.?) (1610)
Total habitat
50 100 150 200 250 300
Dactarge (SWF)
WJa (rIIVIOCCan)
121:03
10.0O3
•
Brown Trout (IFE B2.1 Curves)IALIA (m1/1030m)
3.000
. .
0 0 5 1 1.5 2 2.5 3
aschage (mews)
Figure 135.River Lymington: habitat vs discharge relationships
•EA TechnicalReportWI9
•39
0
5
050 100 I SO 203 250 3C0
Oncharge (WS)
Invertebrates
15
10
5 .
Adult
Ammar,
• Fry
spe5rov
Brown Trout (IFE 82.1 Curves)WJA%(total babb31)
as
30
25
20
IS
10
Mutt
anima°
• Fry
sparang
(Th
Tot nab
WUA (m2/1000m) WUA (% total habitat)
W1JA (rn510D0rn)
1.400
1200
1.000
1300
600
400
200 -
Roach RoachWJA total t atria)
25Mull
-extender 20
0Fry
15
Spawning
10
5-
Mull
Juvenile
Fry
•Spawning
0 01 2 3 4 5 0 103 200 300 400 500
01.1C11enjeCarnal:a) Discharge (%1.16)
WIJA (rn3/100011)
3.500
3.003
2.000 -
1.500
1.000
Dace DaceVMA (%total hetet%)
eoMull
soJuvenile
Fry 43
Spawning 30ice
20
Mull
Juvenile
Fry •
Speowng
500 10
01 2 3 4 5 0 103 200 300 400 503
Drscharge (carnets) Discharge (%131F)
Invertebrates and MacrophyteWUA (m1/1000n6
3.500
3.000 -
2.000 (
1.500
•1003
Invertebrates and MacrophyteWJA Iota Nag)
eo
50
40
o anunat
Levee.
0 Swum.
Lauer.
RanunC.
20 •
500 10 •
0 0 . .1 2 1 4 5 O 103 20() 300 500
Dndiarge (cameo) Discturge (%I•1F)
Total habitat'ANA (mq100Orn)
13000
6.000 •
Tot MO.
4.000
2,000
0 •
1 2 3 4 5
Discharge Cann)
Figure B6. Millstream: habitat vs discharge relationships-
EA Technical ReportW19 40
•
•
•
•
•
•
•
•
••
•••
•
•
•
•
•
•
•
•
•
•
WUA (rn2/1000m)
Brown Trout (Wnsex NRA Curves)Wa 0•1410054
4.003
2 OW
1KO
00345
Crodurpr (ons)
Brown TrOut (WE Et2.1Own)OW11030e0
2303
2 ow
1103
I CCO
00234 5
Ceenewelona)
1/1vOrtObrenWa 001 02001
SOO
4403
3.000
2.003
1,00)
0 0234
Ctch4,74(ono.)
Mac201141Y1o0. 1413A111r1C031
S.=
S.CCO
45033500
2.000
1.003
00I23 4
000 •1011010•011
Total habitaAU* (s*103011120030
15MO
WUA
WUAMa. nabon))
20
25
20
15•
10
5
0
0100
MAC% tail Paten)
20
IS
10
00ICO
WU..col SSC)
35
33
25 •
20
15
10
0
0103
%SAC%Kul mow/53
.0
(% total habitat)
Brown Trout Messax URA Curves)
200303400Deems ClskW)
Brown Trout (ISE B2.1 Cry's)
201300400Cuero Metal
Invertebrate,
•21:0KOCO ChOarge15 1.1F)
Mactollthetos
•203300603Dollop fit MI)
500
503
500
ASS
rytfree•
• SP•4"•11
•
MIA
Frrs
Sean
Sawa
it5Pwring
Gen
Itrry Dew •
RIry.Ca •
•
ftrut
N40.1
fir
Rtessis
Nag 0
20
2,,
10
aoo 100 500
Tam we
•10000
002 3 •
Wen Conga)
Figure B7. River Lambourn: habitat vs discharge relationships
EA Technical Report W19 41
WUA (m2/1000m)
Brown Trout (WE 132.1Curves)WtJA (rn'/1000m)
1 400
1.200
1.000
800
600
100
200
000.511.52
Discharge (cornea)
Adult
Juvenile
• Fry
Spawning
WUA(% total available habitat)
Brown Trout (IFE 62.1 Curves)W1JA % (total Data)
20
Adult
15Juvenile
Fry
910Spawning
5
0050100150200
Drscharge (%P.AF)
Invertebrates and MacrophyteVAJA (mV1000m)
3.000
2.500
2000
1.500
1.000 C 3
500
00.511 52
Discharge (cumecs)
Invertebrates and MacrophyteNAM (% total habitat)
10
30
20r
_110 •
0
010100150200
Discharge (%154F)
Gam mar
Samos.
Wastur.
Gunnar.
Sark.
• Nastur.
Total habitatAIDA (m11000m)
8000
6000
Tot hab
1.000
2.000
00 0.5 1 1.5 2
Discharge (cornea)
Figure B8. Habitat vs Discharge Relationships: River Gwash
EA Technical Report W19 42
•••
APPENDIX C. HABITATTIME SERIES AND DURATION• CURVES
6IM
uoda
ll
1E3!
li4oa
i
V3
Flo
wan
dT
ota
ln
alat
iat
TO
YP
UM
AW
I'N
an)
Flo
wto
nal
Is
IaI
I
; II
I
1I
110
I.I.‘1
1\\r
,-1
\)
I 1)1
Ind
41I,.
•
.4/ t
"l.•
on
,
Ro
ws
WA
YS
ran
ge
Ofm
od
al
*du
dN
UA
f%W
aalM
OW
)
n
l?
cla
•I'1
•el
•
"C
i"C
Iem
u.
I.3
,
I ?a
IIIJ
S.
AA
n
ante
nn
a
MIA
(II
IOW
1111
713)
II
flaw
i/an Ia IS
A Ind
401.
0141
1(,
1:\t•
-..—
ifse
m.
aF
igur
eC
1.1.
Riv
erE
re.
tota
lha
bita
tan
dflo
w,
habi
tat
time
serie
sfo
rbr
own
trou
t(a
dult,
juve
nile
)
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
•
••
••
••
••
••
••
*•
••
••
••
••
••
••
••
••
••
••
EA
Tec
hnic
al
Rep
ort
W19
Y.V
A(%
155
7.7
It
If
11
S.
*nom
.
'4M
a
L
Sp
awn
ing
Flo
wN
WT
)
V&
A(1
4W
assa
y
atia
3W
ITS
DO
WN
WI
MeI
fil•
8V1
tie
Ole
l•v
ltC
IC
O
L„
SsF
igur
eC
I.2.
Riv
erE
xe:h
abit
atti
me
seri
esfo
rbr
own
trou
tQ5y
,spaw
ning
)
au.
(....
..im
etso
EA
Tec
hnic
al
Rep
ort
W19
new
Mar
l
•
I
.140
:
IMJA
(AM
aP
etal
)
;'I
11ri
ait
Ply
\"\
up
aras
A'
nar
(1U
FF
)
I•
a
III
aoI.
Jo,.
\j
t
A_
aF
.cr
,:g
umC
L3.
Riv
erE
xe:
habi
tat
tim
ese
ries
for
inve
rteb
rate
s
asaa
aas
aaas
saIS
a&11
1111
1M
a
River Exe - brown troutFlow Total HabitatArea
3,900
3,700
3,500I
3,300 1
3,100
2,900
2,700
2,500
5 020 30 50 70 80 90 95% of time exceeded
Flow
(%
MF)
Total
Area
00
I()O0n0
11 II I1 15 10 20 30 50 70 80 90 95% of time exceeded
1,000
500
1....11••••1 I I 1200
IGO
50
I I I 11 I 1 *. ------- ...20 J
1 I 1101 1 1 1 1I I I I 11 11
52
••• 11 11 1 1•
Summer Winter
Adult Juvenile
50
WUA
(%
habitat
available)
Spawning
50 . .
11November-
20 December
111
1 I
I I I
I I
5 10 20 30 50 70 80 90 95
% of lime exceeded
1 11 1 1I I I I I I
1 "L
I 1
i I
1 11 1
5 10 20 30 50 70 80 90 9
% of time exceeded
Fry
5 10 20 30 30 70 80 90 95
% of time exceeded
50
10 20 30 50 70 80 90 9% of time exceeded
10
5
2
1
0.5
0.2
0.1
50
30
2 20
tt
1.3 10
-= 5.3%.*
3
2
t4
Figure CI.4. River Exe: Habitat duration curves for brown trout (all life stages) using 16years of daily meanflow data (1966-1981)
EA Technical Report W19 47
River Exe - invertebrates
Summer WinterChloperlidae (A) Leuctridae (A)
I I I
....
10 20 30 50 70 80 90 95
% of time exceeded % of time exceeded
50
T.-•.•
.........
I
20 30 50 70 80 90 95
2
5 10
50
ci 30TSfa 20
toCS
D 3
2
Figure C1.5 . River Exe: Habitat duration curves for invertebratesusing 16years of dailymeanflow data (1966-1981)
EA Technical Report W19 48
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
Flo
wan
dT
ota
lK
ant
(
1 'AS
rn >lo
taS
NO
WW
liCo
Nly
11It
"a
.ses
e
0 PO 70
re
a
.„
II II
11V
i•
skiv
t\sj:v
kihu
t4.\_
.
row
Ser
recr
3. >a
y.eh
.,ci
r411:1
4".,
'11
1.
:IA,
isIt
La..
•
I1,
1III
Sn
C)
no
np
urr
ssw
op
MO
M
Ira
%%
VA
CSs
selr
abir
)
A:s
VI 11
0is
ks‘1
:(11 W•u
vs•
ki
A•
.1Is
rII
o
IIVel
I{ ‘v•s
ofk.
—ss
i'•
:
•
11#11
S "r
Nr.
s•J
• ir
;
jIII1
4{
\.vls
Mau
reu
rsti
o
411
1,il4
As
ser
,‘
VA
SA
ris
tar
hab
lar
3.
Are
na*
Fla
Mae u
se
11
Vkk
,
•I
ki%
\s
—
1q
.S:
‘ivj
••
A--
f4•1
'0•
•._
kkA
in"
.1/2
,•.1
)I#‘
1Ij
•In
L4'`,
111
‘ks\
J"ks
h`4‘
,y 1.1s
Ad.
)
,I
VJI
a•
I‘ssi
VS
\:4
Iik
&t•
P•
k•rt
-
1/4eF
igur
eC
2.1.
Riv
erW
ye:t
otal
habi
tata
ndfl
ow,
habi
tatt
ime
seri
esfo
rbr
own
trou
t(ad
ult,j
uven
ile)
WJA
Mal
MU
M)
new
(IM
F)
Fry
MIA
(II h
aISM
ew)
Fla
.()
Yr
I
11.rlo
k:‘,_
111
1 CI.'
kk,'
• 'II.
Cl
v ,4
\\NP
11I
11.
Ia
i.i
it‘
I14
1I
OS
1.1%
11.
15
Lh
e4:
1,
h,•
'
I•14
,;\,
\P
Pcu
LA.,4
\.;\
U',A
d1
070
to
Ln
Fig
ure
C2.
2.R
iver
Wye
:ha
bita
ttim
ese
ries
for
brow
ntr
out
(fly
and
spaw
ning
)0 •
••
••
•• •
• •• •
• ••
••
•• •
•• •
• ••
••
••
••
•
EA
Tec
hnic
al
Rep
ort
W19
*vile
Irtr
i In
s
L•N
j4..
• th.
.i ,
a
vi
it.
11
\')j'e
N.
OV
3141
101
IC
I
I,'I:
I•
11
•I
;v
'
\\N
KO
NO
•s I
,5
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
t 4./
IIS
:/‘
4V
•V\
erve
rr.
ftnumwto.
NtskOkkan.4140
rra
'1/4
,I jikJ.
)V.
;!
,i;14
•I;I
I•al.'111)1
i\j/
rd\,s
,,
.10
i
1,4;1
ir!y
1,
`Min
e
i.e
lan o,41
5
" \so
\.;
1/2;
%
I
;A..)
IIr•
61m
poda
H
po!u
tpol
va
lare
skin
4)4
h11 .1
.1
,,
up
eole
IK•N
•arn
t-^
Fig
ureC
2.3.
Riv
erW
ye:h
abit
atti
mes
erie
sfor
inve
rteb
rate
s,plu
sran
uncu
lus
Fp
.11
6m0)
Ia
In
PO
Wea
lrap
aRIVOTOICulaa
II
kj,j
:1/2
1<51
5. ‘11
_4 \1.1
\2:L
,Lu
i'424
._;..
,na
l
111,
L011
,1.
)\k/
iiI\
VI
Flo
w0v
A.)
!er
EA Technical Report W19 52
Flow
River Wye - browntroutTotalHabitat Arca
1,000
500
200
100
2 50
g 20
10
5
2
11 1 1
1 11 11
1 11 1I
I ! I I
III I!
11 1
1.%".
----......;,...........
Ii •-:..1 .
Total
Area
(m'/1000m)
16,000
14,000
12,000
10,000
5 10 20 30 50 70 80 90 958,000
% of time exceeded
Adult
1 1 1
1 1 1
$ 0 20 30 50 70 80 90 95% of timeexceeded
50
1
Spawning
SummerWinter
50
30.
20'1
10.
5
3
2
1
% of
I
time
-
I
Juvenile
exceeded
I iIII•
I
= ..c=Cd
e'
<C
I I
.10 20 30 50 30.80 90 95
V. of time exceeded
Fry
Figure C2.4.. River Wye. Habitat duration curvesfor browntrout using-16years of daily meanflow data (1966-1981)
50
302020
WUA
(%habitat
available)
WUA
(%habitat
available) 10
5
3
2
1
5 10 20 30 50 70 80 90 95
% of time exceeded
I I 1I I
I II I I11
50November-December
10
5
2
1
0.5
0.2
0.15 10 20 30 50 70 80 90 95
% of timeexceeded
River Wye - invertebrates and macrophytes
Summer Winter
Leuctridae (A)
WUA
(%
habitat
available)
50
30
20
10
5
3WUA
(%
habitat
available)
50
30
20
10
5
3
2 2
Polycentropus flavomaculatus
I I I II I I
I I I
15 10 20 30 50 70 80 90 95 5 10 20 30 SO 70 80 90 95
50
I30
= 20.0
"l10
5
3
2
II
% of
I I 1
time
Ranunculus
exceeded
I
I1
I
I
% of time cxceeded
5 10 20 30 50 70 80 90 95
% of time exceeded
Figure C2.5. River Wye: habitat duration curves for invertebrates and macrophyte using 16years of daily meanflow data (1966-1981)
EA Technical Report W19 53
EA
Tec
hnic
al
Rep
ort
W19T
awS
ala
WI•
IWY
,/
wit
*
no
wan
dT
ota
l MA
NIA
tA
to)
\A
,.
•M
aWel
RO
WSC
U01
9039
1M
el
Mal
VO
AM
tn&
NO
W
•
JOvo
n11
41
no
w(I
IM
a)
XI
IV
W
5(
lad
5I
ij1
'V
. I-A
: C -;\
----
`i\-
1`--
----
A--
--
1-
1\/
1til
\C
a
p.A
.J.J
1\.—
.....n
.fr•-
-1LA
—._
_-_)\
----
• mu.
saw
n...
...,
Fig
ure
C3.
I.R
iver
Bli
the:
tota
l hab
itat
andf
low
,ha
bita
t tim
ese
ries
for
dace
(adu
ltan
djuv
enil
e)
wuA
floK
ulm
ow
0021
/10
03 1
141
),
1A
1111
.1
'••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
VO
JA(e
.M
el/W
illi)
(%40
)
I07
6IpA
poda
N
iBoi
tiqoa
l
Vg ID
1.;
A.1
1,1
jVJl
‘t-
-
0001
Sp
awn
ing
VA
JA(l
bW
Mb
enig
)F
lom
I-%
WI
IK
O
Fig
ure
C3.
2.R
iver
Bli
the.
Hab
itat
tim
ese
ries
for
dace
(fry
and
spaw
ning
)
WIJ
A(1
101a
lhab
l103
"M
AC
MA
JAt%
gala
nsem
a0
Mal
001-
1021
01M
A.•
•na.
6IM
voda
ll
luol
utlo
a1
va
Pk.
(Ire
Of) C
M
Xia IM
O
••••
'"
I ,In
.,1
.I
II
--
SS
.*@
Min
i
Fku
reC
3.3.
Riv
erB
lithe
:ha
bita
ttim
ese
ries
for
roac
h(a
dult
and
juve
nile
)
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
Frr
,W
U/.
W30
1111
tab
400
Flo
w(I
IM
r)
2a
IX
0
EA
Tec
hnic
al
Rep
ort
W19
*IP
XP
WrI
ng
WJA
(.4
MIS
Wh
in)
Po
w1%
kin
II$
It
0$ %
PM
*•~
1P
lort
l0~
10'v
ow
Co
row
VP
.01
.01
MIX
7,90
0.0
IP
u,
Fig
ure
C3.
4.R
iver
Bli
the:
habi
tatt
ime
seri
esfo
rro
ach
(fry
and
spaw
ning
)
Oat
inkl
as
61N
s.
uoci
aN
pop.
npa
1Vg
V&
A(A
toll
MC
I&•
Olc
oo
rML
IF)
zota
a•
,,:j\
j,„J.
);.,\
„;tv
) te„L
-,;n
, 41
1411
Pliw
ownw
l.•••
•%
WA
r111
041
hat
0.13
coR
gure
C3.
5.R
iver
Bli
the:
habi
tat t
ime
seri
esfo
rtw
oge
nera
ofin
vert
ebra
tes
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
•
River Blithe - daceFlow TotalHabitatArca
1,000 13,000
• 1 1
s% I 12,0001 1
1 1
1 1 1 1
1 1 1 1
500 •••*"
200
100
2 50
20
10
5
2
50
-------:I ..,,
20
10
5
O
:h.2
3
e 1
<n 3
0.5
0.2
0.1
1I ,000
10,000T
I H9,000
% of time exceeded % of time exceeded
Summer Winter
Adult Juvenile. 50
- 1l
I
20.....
1 =0, 1O 5g 1lran 2
.. 1 .=e I
t- 1 < 0.5tl 31N. 0.2
% of time exceeded
I1,...... _L...7„„,I0
it0.110 20 30 50 70 80 90 95
%of time exceeded
Spawning Fry
50
30
20
>C3
10
C3
5
3
2
I I1 1 1 1
March-April
50
30
—920
C3 1017:1
5
a 3
2
5 10 20 30 50 7 80 90 95 5 10 20 30 50 70 80 90 95
% of time exceeded % of time exceeded
Figure C3.6. River Blithe: habitat duration curvesfor dace using 16years ofdaily mean flowdata (1966-1981)
EA Technical Report W19 59
WUA
(%
habitat
available)
50
30I
20
10
5
3
2I
2
1
0.5
0.3
0.2
0.1
I5
510
1
10%
20
20of
IHI
Spawning
30
Adult
3050 time
50
exceeded
70
70
1
1
80
April-June
90
I1
1I
1;95
River Blithe
Summer
- roach
Winter
50
30
=,20"a
10
Ct.o5
<3
2
1
50
30
= 20.0Ct
Ct
II107.71ISCt
<3
2
5
11
5
1
1
1:1
80 90 95
% of time exceeded
Juvenile
I
I '1 'I
i iI I
10 20 30 50 70 80 90 95% of time exceeded
10 20 30 50 70 80 90 95
% of time exceeded
Fry
Figure C3.7. River Blithe: habitat duration curvesfor roach using 16years of daily meanflow data (1966-1981)
...
EA Technical Report W19 60
River Blithe - invertebrates and macrophytes
Summer WinterPolycentropur flavornat Wants Gammaridae (abundance)
50
30
.013 20
.0> 10
.c $
<
2
1
50
30
.01: 20
ea> o3it.c 5
2
1
50
30
.01: 20
"ts 10A-1g.= 5
<
2
1
Nasturtium
10 20 30 50 70 80 90 95
5 10 20 30 50 70 80 90 95% of time exceeded
% of time exceeded
% of time exceeded
Figure C3.8. Habitat duration curves for invertebrates using 16years of daily meanflow data (1966-1981)
EA Technical Report W19 61
Fle
ear
sT
em+
mab
litat
MO
She
liAM
to*,
NO
W)
now
Mew
s)11
030
6IM
Inda
)1
leo!
ullo
al
vg
XX
IO
.O
xlm
ers
Xha
Mal
t,IN
A.
Ira
ga"
••M
s
AA
Atr
ait
RA
NA
Flo
wM
IS) X
O
I
cssi
v.)
Fig
ure
C4.
1.R
iver
liche
n:to
tal
habi
tat
and
flow
.ha
bita
ttim
ese
ries
for
adul
tbr
own
trou
tct
%
,a.
aaa
aa
&m
amm
as-m
ama
aan
a...
..a
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
Fry
/Jo
ry•F
a•W
IJA
alto
n!in
t4•0
ins
(%IF
F)
ano
61Ak
uoda
u
lupp
pai,
lia
X0 IP
r"A
N
!
Ova
.av
a•o•
•linn
s*a
wn
•un
etto
r,ne
in)&
nun
swan
s•ar
n
Fpr
anng
W
UA
(%la
aN
NW
)n
onLa
ir) ?IQ
Don
eW
aal
WIN
DW
WI
wan
man
Win
ola
wn
&nu
n&
nue
flan/
e`fi
gure
C4.
2.R
iver
Itch
en:h
abit
atti
me
seri
esfo
rbr
own
trou
t (fr
y/ju
veni
lean
dsp
awni
ng)
1/0
Oan
efus
Pu
ler
V&
Ala
alM
ote
)
Oa
61PA
uoda
ll
le3P-a1
301V3
no
w(I
MF
) re
.••
a
Mai
ne
rain
s
000) 1I:
••••
••1
••••
••e.
t.a
Isn
.M
OM
.••
••••
,11
LPM
nWfl
da
(Al
WU
AM
I11
1M14
WIP
1411
1S
te—
Vie
n.4-
1.01
173
•1•1
4101
.140
@P
IUS
\AP
''14
.7;Fig
ure
C4.
3.R
iver
liche
n:to
tal
habi
tat
and
flow
,ha
bita
ttim
ese
ries
for
two
gene
raof
inve
rteb
rate
s
alan
aal
ban
/11
III•1
11d
bill
ita•
1111
allt
allh
ellh
adlia
aaal
lIa
Flow Total HabitatArea
200
100u":
1,000
R iver Itchen - brown trout
500
l I 1 'g 20
10
2
I I I5
20,000
—...19000E
5; 18,000
17,000t-
5
50
30
:On20r_- reei I0an
;tit! 5
D 33
2
16,000
% of time exceeded
I
170 80 90
SummerWinter
50
20
10
5
2
1
0.5
0.2
0.1
n-7
I
•
16
A'
95
Spawning
November-1December
10 20 30 50% of time exceeded
50
30
Adult
WUA
(%
habitat
available)
20
10
5
3
2
% of time exceeded
Fry /Juvenile
I I
I
% of time exceeded
5 10 20 30 50 70 80 90 95
% of time exceeded
Figure C4.4. River lichen: habitat duration curvesfor brown trout using 16years of dailymean flow data (1966-1981)
EA Technical Report W19 65
River Itchen - invertebrates and macrophytes
Ephemeridac(A)Summer Winter
Ganintarus pules
50 50 r
— 30
es 20C3
I
3
2
30
20
10
5
3WUA
(%
habitat
available)
2;
5 10 20 30 50 70 80 90 95% of time exceeded
% of time exceeded
Ranunculus Nasturtium
WUA
(%
habitat
available)
50
30
20
10
5
3
2
I I30I I I20
11.
10
5
3
50
WUA
(%
habitat
available)
5 10 20 30 50 70 80 90 95
% of time exceeded5 10 20 30 50 70 80 90 95
% of time exceeded
Figure C4.5. River Itchen: habitat duration curves for invertebrates and mocrophytes using16 years of daily mean flow data (1966-1981) -
EA Technical Report W19 66
••••
••••
••••
••••
••••
••••
••••
•sew
ee
61pA
podo
s
leo!
uqoa
l,
vgL
OS
Pat
illw
enm
oo11
Pim
alT
otal
•601
91F
low
Ow
es)
•
0
I/1
I/A
II
I(to
,,t
0•/
II.It
)••
sT
ilIk L
i
A•-
_,.,
L.,
\I4
\,,,A
JQ
V
0,4.
I!I
I!I
I'
II,I
)
::it
1I,
,.4:l
iill'P
1;''
I.
I:It
,i,6t
,h
..?
II)?I
'll,:
:II;
l'ivi
;\ti.
4Is'
:10
v1A
yvv
vj\',
N'
0071
,
•
hip
S1::
!,
1.4
•
Fla
ws
outs
acle
rang
eet
WA
rase
ssaw
new
rAS
in
t;
AA
;SA
AA
,
A
I.•
1/4,-
,-;\A
.,'
703
•
I,•
,,
' \j/
MIA
(14
te4a
11.0
014
lvto
na.
71e1
1no
wrs
usi
I,S
ik_A
P•‘\
JO
LPA
:1/4
-.
Fig
ure
C5.
1.R
iver
Lym
ingt
on:
tota
lhab
itata
ndflo
w,
habi
tatt
ime
serie
sfor
brow
ntr
out
(adu
lt,ju
veni
le)
ON
0,/
EA
Tec
hnic
al
Rep
ort
W19 V
elM
eler
e./
esM
e,
.tn
elro
leF
ie••
••••
••••
••e
rnS
ee.
ea,
Fig
ure
C5.
2.R
iver
Lym
ingt
on:
habi
tat
tim
ese
ries
for
brow
ntr
out
(fly
,sp
awni
ng)
00
••••
•••
••••
••
••
••"
••
••
••
••
••
••
••
••
•
•••
••••
•II•
•••
••••
•••
••
••••
••••
••
••
WU
AIO
Wn
ud
e)b
ren
viti
a•1A
Irm
ar.
61m
,
uoda
a
ro!u
tpa±
vg
i;is
IIII
.
I"
"t:
,!O
.•
ij
I
•
11
I•l
•IV
!
I
Sn
••••
,I:
Lat
rId
Oe1
(A)
wu
*to
w/m
eta)
,I
0
It
:1:
.1:i
(ie
: 1r
iIil
•1
.IC
.:;
1?.
is.
1II,
I0,
,1;1
111
•li
•ri
i
'.
.
,.
li1/
41,':
;:,,P
,,..,
m,
,,,,
„,i
,.
I,'j
4•_
.;1„
.1j1
41,‘,
,I1,1
/4.,:
vi,
j,,
11
1t
.I
.,.
km'y
i•
I1
"\
t•-!
1/4.
....•\
•‘..
4\1
`‘)\\,
,,i,,,
,,\
1.•\
•••
•••-
••—n,
AA
_
Mo
t(V
AA
F)
ao,,p
I
14I,
••
i?i
1,1
'?S
IP,J
OIN
?'P
'IS
4
r.11
Y1
•I
I",.•
1,
\4,
1
01.0
1411
MS
SIn
AV
711
SF
igur
eC
5.3.
Riv
erL
ymin
gton
.Hab
itat
tim
ese
ries
for
two
gene
raof
inve
rteb
rate
s
River Lymington - brown troutFlow TotalHabitatArea
1,000 11,000
I% 1 1111I I I50
I : I 7,000
I I I-11
6,000
500
200
100
10,000
9,000
8,000
% of time exceeded % of time exceeded
Adult
1 I I
I I I
.1%
N.
,
0.15 10 20 30 50 70 80 90 95
% of time exceeded
I November-t I
I I IDecember
I I I
I I
0110 0 30 50 70 80 90 95
% of time exceeded
0.15 10 20 30 50 70 80 90 95
Summer WinterJuvenile
50
20
10
5
2
0.5
02
0.15 10 20 30 50 70 80 90 95
WUA
(%
habitat
available)
Spawning Fry
50 50
WUA
(%
habitat
available)
20
ID E
5 '
2
1
0.5
0.2
WUA
(%
habitat
available)
20
10
5
2
50
20
10
5
2
1
0.5
0.2
WUA
(%
habitat
available)
20
10
5
2
1
0 5
0.2
% of timc exceeded % of time exceeded
Figure C5.4. River Lymington. habitat duration curves for brown trout using 16years of dailymean flow data (1966-1981)
EA Technical Report WI9 70
River Lymington - invertebrates and macrophytes
50
30.0co 20"ra
10
e 5
3
2
5 0 20 0 50 7080 90 95
% of time exceeded
Nasturtium
30
20
10
5
3
2
1
Summer Winter
Leumidae(A)
, .
. .I -
II
I I I I \
I I% of time exceeded
Ranunculus
Gammaridae (A)
I I50
WUA
(%
habitat
available)
50
30r,z° 20It
fl
'I
la
10
.c 5
a 3
2
50
30
20
10
5
3
2
15 10 20 30 50 70 80 90 95 5 10 20 30 50 70 80 90 95
% of time exceeded % of time exceeded
III
WUA
(%
habitat
available)
Figure C5.5. River Lymington: habitat duration curvesfor invertebrates/ macrophytesusing 16years of daily meanflow data (1966-1981)
EA Technical Report W19 71
61A
k
poda
li
reom
ogi
Nig
Taa
lhab
at(S
1100
0m)
*NIS
8701
4700
46
Far
and
Tea
l/4
6/41
I.
r ;1%4
ttt
s•
61)?)I „
\ \
IS
AA
hi
AO
AL
461:
%/L
I'w
•na
ltMel
„
MIT
WI
sow
n
Flo
wso
utsi
d•ra
no,a
tm
odal
ow
nM
AO
*M
in
• r 114'
.1`4
/1.4
0,
'
\
'N
jkl,
h•di
WJA
WM
SO
W)
•
Ad
a
Nh
14, ik 4ri
;A\
P
1:1 \41
• 0'
It-
WIM
P
,
0941
8.1
001/
//11/
0014
1
I
\1
\0.
11.4
11O
S14
1W
ang
Vi1
IW
/7C
•01/
7
J••
••11
6
WI
WJA
Isto
1hai
slel
)
OS
A/
V41
44
PO
O.
--.1 I.)
Fig
ure
C6.
I.M
illst
ream
:to
tal
habi
tat
and
flow
.H
abita
ttim
ese
ries
for
dace
(adu
ltan
dju
veni
le)
•• •
••
•• •
••
••
••
•• •
••••
••••
• ••
••
••
•
••
••
••
0••
••••
••••
•0••
••••
•••
•••
•••
EA
Tec
hnic
al
Rep
ort
WI9
INIO
• ••
•• •
••••
••1
,m
44v.
\11
,
11.1
.•••
•.
eJai
fre
W7I
64-1
:101
Pe-
0101
1
Pla
,(1.
1/1
Ftg
ure
C62
.M
ills
trea
m:h
abit
atti
me
seri
esfo
rda
ce(f
iyan
dsp
awni
ng)
4.;
1Y
AVI
i 00.0
)
\01
1131
•7I.1,0k:
(,)!
':.
-J\
en.
ik
NM
.
.1
1IY
l;el
1'L
i51
11 wvI
it...1
•_a
tm
.
o. io4A
‘VA
it ''.7
k\54
515/
,'il-1
11
• Njo
ysto
.leiu
i•
+eh „
If ;I
.1.
III
re.t
eas$
laT
OT
.
I•to
PO
I
6IA
\
igot
utpa
l,
Nig
/am
ino
@S
nei
ram
lF
..T
ealIt
ed
tO,'re
.I
1.41
11rk
ki,
\.0,4
0v7s
AY
R04
w$
eliim
elsw
am
NM
,N
an
10 •
••
••••
I .0.4
1100
000
0
•"4
v Os%
).1
1l`
k
M10
12N
OU
N10
0
aae
,
Fig
ure
C6.
3.M
ills
trea
m:t
otal
habi
tata
ndfl
ow,
habi
tatt
ime
seri
esfo
rro
ach
(adu
ltan
djuv
enil
e)
.11
1111
1111
111a
ailld
hai
ltaa
aaaa
aaa.
aaaa
daa
aaaa
.....
••••
••
0••
••
••
••
••
••
••••
••
••
••
••
••
•
61M
vocl
o-H
lua!
utpo
i
V3
k.
WW
11/
401/
01co
ovn
i„
Fry
H.„
't0/4
t.60
01.6
7C
OID
la04
01.0
0W
2V71
CO
ON
,
.vA
jsA
NA
5040
0110
001
4011
4as
Sla
y10
0.1%
/40
SC
O1
4F
low
1240
05.
0a0
1,
Viw
IIn
0.I
I‘T
h..)
08 0• 04
021.
1-1•
0V
Ma
I.
\\0(
,10
0
00 01
400
0/04
01C
lan
0LO
V01
01.0
1/70
0190
/11
0034
0701
.340
1&
LO
VA
4101
473
•1.0
0111
01a/
1701
0411
I•a
n,
Ote
a01
0044
1
--I
Fig
ure
C6.
4.M
ills
trea
m:h
abit
atti
me
seri
esfo
rro
ach
(fry
and
spaw
ning
)1.
11
Flo
w
1°.
ble
r
AIM
VA
SA
Mb
tn.
ito
blo
g
.914
1,41
.O
gle
rn
ow
.C
09.•
CS
OU
7W
ON
/C
eelf
re
EA
Tec
hnic
al
Rep
ort
W19
IH
./4 1/4„
,,
MO
W.
0014
4
NU
A(1
1IO
WK
ON
A
0011
1111
11O
M(.
1
Leu
clA
da
(A)
arev
n
no
w(%
10) R
fe
LR
CM
-
".1
h
cr,
Fig
ure
C6.
5.M
illst
reat
n:ha
bita
ttim
ese
ries
for
two
gene
raof
inve
rteb
rate
s
Millstream - daceFlow Total HabitatArea
0% of time exceeded % of time exceeded
Summer WinterAdult Juvenile
2
5,000
11
11
Total
Area
(rn'Il000m)
7,000
6,500
6,000
5,500
200
100
50
220
10
5
50 50
30
St 20
. a>10
5
3
30
20
io
5
< =22
15 10
\
20 30 50 70 80 90 951
% of time exceeded
10 20 30 50 70 80 90 95% of timc exceeded
5050
3030
2020
10
5
l o
3
2WUA
(%
habitat
available)
WUA
(%
habitat
available)
190 95
Fry
5 10 20 0 50 70 80 90 95
Spawning
I
HI I
March-April
5
3
1
2
1
!
510230570 80
% of time exceeded % of time exceeded
Figure C6.6. Millstream: habitat duration curvesfor dace using 16years of daily mean flowdata (1966-1981)
EA Technical Report W19 77
Millstream - roachFlow Total Habitat Area
5
I
I1
200
100
50
20
2
I
I
I
I
III I1%.
EItI I0
III
7,000
6,500
6,000
5,500
5,000
%.
510
20 30
I
50
1
70 80 90 95% of time exceeded %of time exceeded
Adult
I I
1, I II
I I I—
I I1
5 10 20 30 50 70 80 90 95 % of time exceeded
Summer Winter
50
30
20
10
5
3
2
1
Juvenile
% of time exceeded
WUA
(%
habitat
available)
50
30
20
10
5
3
2
1
WUA
(
habitat
available)
2
0.1
Spawning Fry
! 50
April-June
5 0 20 30 50 70 80 90 95 5 10 20 30
I II 1
11
1 1
50 70 80 90 95
30
117 20.0.ta>m 10
.0 5
< 3
' 2
% of time exceeded % of time exceeded
Figure C6.7. Millstream: habitat duration curvesfor roach using 16years of daily meanflow data (1966-1981)
EA Technical Report WI9 78
Millstream - invertebrates and macrophytes
Flow Total 11abttatArea
% of time exceeded% of time exceeded
I II II I I II I
\
% of time exceeded
SummerLeuctridac (A)
90 95
7,000
6,500
C
E 6,000Cl
Toot5,500
5,000
Winter
50
30I
2 20.0
n 10
5
73
2
510%
I
of
I
20
time
Gammaridac
30 5070
exceeded
(A)
*.
80
N
I
90
I
95
200
100
50
2
20
Is- to
5
2
50
30
1, 20
10
-g5
3
2
ICI 20 30 50 70 80
Ranunculus
50
30
I 20co
> cu
1Eco.c
10
5
3
2
10 20 30 50 70 80 90 95
% of time exceeded
Figure C6.8. Millstream: habitat duration curves using 16years of daily meanflow data(1966-1981)
EA Technical Report W19 79
EA
Tec
hnic
al
Rep
ort
W19
CY
-al
To
tal
Hal
al
• tat
•el
••••
1011
.
0141
1.0.
141
41•1
1
flys
(an
.)
•len
]O
U01
17.1
MN
*M
ein
).0
1 71
Min
Flo
ws
outs
ide
rive
egm
ois
00F
igur
eC
7.I.
Riv
erLa
mbo
urn:
tota
lha
bita
tan
dflo
w,
habi
tat
lime
serie
sfo
rad
ult
brow
ntr
out
••••
••••
••••
••••
•flm
ense
ssam
a
••
••
••
0•
••
••
••
••
••
••
••
••
••
••
••
•••
•
61m
uoda
8
iga!
utpa
i
vg
reA
nIn
00
VA
SA
rit
lols
So
w)
10,4
1011
11sa
gas
ON
ISS
IDM
AO
SIA
M
1
ISS
/71,
0;04
INO
SIP
SIM
AM
WU
AM
YM
UM
)F
lew
Min
no
Pa
inte
aP
ala
OvI
VI1
eVIN
I•u
va•
eiss
aa!W
irt,
Fig
ure
C7.
2.
Riv
erLa
mbo
urn:
habi
tat
time
serie
sfo
rbr
own
trou
t(f
ry/
juve
nile
&sp
awni
ng)
6IPA
uoda
ll
leoP
-
11-133.1
.V
H NA
Nel
bIN
NIA
JR
hoca
raN
aaw
ait
Fla
NA
N)
WN
Wne
ws fo
ibr
agm
ow
tad
IN
em
Gan
enal
dN
IA)
SW
UM
•••
71
MN
.N
ON
*
Fb
.(%
IS)
-1 INN
S.9
1.41
11Ilt
4tff
I*Y
in.
0~1
60,4
1
gl)3F
igur
eC
7.3.
Riv
erLa
mbo
urn:
habi
tat
time
serie
sfo
rtw
oge
nera
ofin
vert
ebra
tes
'••••
•••
••••
•••
•••
,adI
llesa
nsas
ass
ma
RiverLambourn- browntroutFlow TotalHabitatArca
1,000• 500
200100
2 50• 320• 10
5•
2•
15,000
14,000
13,000
Total
Area
(nf/l000m)
12,000
11,000
•
•
•
•
•
••
•
•
•
•
•
•
•
•
•
•
•
:E° 7703>
zn=<
c;rz.
.2
=
5030 20
10
5
32
1
50
30 20
10
5
32
1
•
I
1 5
.•5
1020 305070 8090% of timc exceeded
Adult
-- 7-
! I
I11020 305070 8090
- % of time exceeded
Spawning
November-December
95
95
I
95
SummerWinter
5
50
20 105
2 1
0.5
0.2 0.1
10
510
20 305070 80% of time exceeded
Fry /Ju‘enile
11•
20305070 80
% of time exceeded
90
90
95
9
IC=
1020 305070 80
11
90
% of time exceeded
Figure C7.4. River Lambourn, habitat duration curvesfor brown trout using 16years of dailymeanflow data (1966-1981)
EA Technical Report W19 83
River Lambourn - invertebrates and macrophytes
Summer WinterGammaridae (A) Rhyacophilia dorsalis
50 50
sr; 30.0
NTtse3>
.3 10
-ce 5
= 3
2
I I I .0 20 30 50 70 80 90 95% of time exceeded
5
2
5 10 20 30 50
I
70 80 90 95
% of time exceeded
Ranunculus Nastunium
I I70 80 90 95
WUA
(%
habitat
available)
30
50
20
10
5
2
3
WUA
(%
habitat
available)
30
20
50
10
3
2
5 10 2030 5010 20 30 50 70 80 90 95
% of time exceeded % of time exceeded
Figure C7.5. River Lambourn, habitat duration curves for invertebrates and macrophytesusing 16years of daily mean flow data (1966-1981) . _ . .
EA Technical Rcport W19 84
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
law
adto
ter
Ilab
rait
,\
PO
V•0
61M
uoda
ll
IE3!m1
0oi.va
Tar
rR
obot
(WA
Wm
)
MO
Mse
a
new
(wen
n) Ia
s,•
Ada
Tro
utW
SWM
IA(1
4ro
tsru
erm
arr
Flo
wW
arn
saw
' ',
,W„ w
:-
\M
NI
\\
\V
i
• 1/4
,1V
1..1
/4;
Ska
la
flog
sovu
leru
gsci
f
Jan
alb
Tra
ilK
rbit
s*
WU
A%
1dM
hab1
21)
Fle
aM
ille)
1C1 -,;
kn,
\P.
,\.,
,
Vr
arc
cre
Fig
ure
C8.
1.R
iver
Gw
ash
atB
elm
esth
orpe
.Tot
alha
bita
tand
flow
.H
abit
atti
me
seri
esfo
rbr
own
trou
t(ad
ult,j
uven
ile)
Fry
Tr.
:NIM
OW
wuA
risla
mre
rile
•
Mo
oow
n Ne
EA
Tec
hnic
al
Rep
ort
W19
a1
Ia4
•
ie;1
Ikle
oe
I
lin,\-
-Aiv
g\.
1•1
4.••
• ralW
aallu
bta
llS
eala
nn
eT
ata
WA
Y
Ikm
,p
ad, I
Xe
WI*
,.•
•
Ilan
LeO
VIV
N
,
.•
AIM
OW
xi*V
I07
1NA
V?
.10I
071M
inn
,Cio
Nn
eCN
N.*
1011
41
acgF
igur
eC
8.2.
Riv
erG
was
hat
Bel
mes
thor
pe.T
otal
habi
tata
ndfl
ow.
Hab
itat
tim
ese
ries
for
brow
ntr
out(
fry,
spaw
ning
)
••••
••••
••••
••••
•••
ass
sass
saA
saas
/
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
•
61m
2.10
dall
p011
.114
001
V3
Gro
rld
a.(A
)
/ ill
:11
—4
ArP
lafj
,k
1/41
.I
Sed
solo
wal
ida
1.41
,
VA
L/I
ist,
/A
SS
aWU
AS
Wh
ob
itsp
as 3.
;
jr-
vAJ\it
;;
ivy
N'
)e
,j11
I,V
44
ft4P
UN
Ire
in,
no
wIS
4M/I
'ow
3.
.41,
‘P
,;is
4tIt
;
D ?,),
1
N
Vis
N.
1\jy
Nv/
121
\,
1.,
n`.
..1
, .,
v/ .)
;l:w
'
1;
1i
Licr
t\aP
J1
l\ol
3/
GS
M
coF
igur
eC
8.3.
Riv
erG
was
hat
Bel
mes
thor
pe:
tota
lha
bita
tan
dflo
w,
habi
tat
time
serie
sfo
rin
vert
ebra
tes
River Gwash - brown troutFlow
I 1III 8,000
' I II
7,500I
8
'h 7,000
3
I
1 -26'500
6,000
% of timeexceeded
Total Habitat Arca
5 10 20 30 50 70 80 90 95% of timeexceeded
1,000
500
200
100
g 20
10
5
2
510
1
1H
%20
II
of30
Spawning
Adult
1
time
1
50 70exceeded
November-December
1
1 11111
8090
rs
95
Summer Winter
50
20
71:110
5
AZ2.c
1
0.5
0.2
0.1
50
20tl
0
5
:13CO 2
e
0.5
0.2
0 1
1
III
1
10 20 30 50 70 80 90 95% of timeexceeded
50
20
10
5
WUA
(%
habitat
available)
2
1
0 5
0.2
0.1
0.01
Juvenile
I5 ICI 20 30 50 70 80 90 95
% of timeexceededFry
I I
I I
I I
1. I 1
15 10 20 30 50 70 80 90 95
% of timeexceeded
WUA
(%
habitat
available)
30
10
3
1
0.3
0.1
0.03
Figure C8.4. River Gwash: habitat duration curves using 16years of daily mean flow data(1966-1981)
EA Technical Report W19 88
River Gwash - invertebrates and macrophytes
50
3073es•20
'a10IS..cZit,'5
3
2
Gammaridae
I
i
I
1
(A)
I
1
SummerWinter
I
I
II;
Sericostomatidae
I
(A)
I I
II
50
ci 30
<I20
"i to
3
2
II10 2030 50 70 80 90 95% of time exceeded
5 5 0 20 30 50 70 80 90 95
% of time exceeded
Nasturtium
50
5 10 2030 50 70 80 90 9% of time exceeded
Figure C8.5. River Gwash: habitat duration curves for invertebrates and macrophytesusing 16 years of daily mean flow data (1966-1981)
EA Technical Report W19 89
APPENDIX D. FURTHER ANALYSIS: MISSING VALUES,ADDING FLOW TO A HABITAT DURATION CURVE,SINGLE-YEAR DURATION CURVES AND FLOW &HABITAT SPELL ANALYSIS
...
90
•••
RiverLymington,Browntroutfry••
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
t o3..152
.
..F33
ii c%
<
40
30
20
10—9 8765
4
3
21
--,
2
—.--,
5
_.-,
10
---•
20 30 50
A
70 80 90 95 98 99
300
200
——10900BO 70
50030
.20
91086543
2
M5*0Km
% exceedance
Figure Ala. River Lytnington: habitat duration curve withflows correspondingto habitat alsoplottedNote that thisdoesNOTshowflow erceedancepercentdes.
•
•
•
•
•
•
•
•
EATechnicalReportW19 91
•
•
40
30
20.
River Lymington, brown trout fry
300
200
- 10900 41
80
70
60
50
co
40
co
.c
10 30
_to9
820
70
< 8
5 - 109
8
4 7
5
3 4 •
3
•2 2
1 2 5 10 20 a xcadara 80 90 95 98 99
Figure D.I b. River Lymington: flow duration curve wIth
corresponding habitat values for brown trout fry
EA Technical Report W19
River Lymington, Nasturtium40 300
20030
•20 _100 90BO
• • • 2 w10
7060500
30
• z9
#.3
• ci87
20-n
• < 6
• 5
910
8
• 4
7
6
•
5
3
•
3
•
2
2
•
1 2 5 1020 305070 8090 9598 99
% exceedance
•
Figure D.2a. River Lymington: habitat duration curve with corresponding flows for NasturtiumNote that this shows habitat, exceedance percentiles, not flow exceedance percentiles.
•
• EA Technical Report W 19 93
•
WUA
(%
available
habitat)
40
River Lymington, Nasturtium
300
30
200
20
100
90
80
70
60
50
10
30 - n
9
8
20 7,C8
7
6
10
9
5
8
•
7
6
•4
5
•3
3
•
2
•2
1 2 5 1020305070 80 90 9598 99
•
% exceedance
•
Figure D.2b. River Lymington: flow duration curve with
corresponding habitat values for Nasturtium
EA Technical Report WI9
River Lymington, Ranunculus40 300
30 200
20 100 9080
•
•
••
Car
.210Fa 9
o 843>
7:*
< 6
5
4
3
21 2 5 1020305070
%exceedance8090 9598 99
0
6050
0
30
20
109
.7
6 5
4
3
2
0-
-T1
Figure 113. River Lymington: habitat duration curve with correspondingflowsfor Ranunculus
Note that this shows habitat exceedance percentiles, not flow exceedance percentiles.•
•
EATechnicalReportW19 95
•
•
River Lymington, Ranunculus40 300
30 200
100so so70
60
50
40
30
20
10. 9. 87
5
3
2 21 2 5 10 20 30 50 70 80 90 95 98 99
% exceedance
Figure D.3b. River Lymington: flow duration curve with
corresponding habitat values for Ranunculus
20
4
3
EA Technical Report W19 96
Ii • :\\r
t\.4
pr
rif9
.fnin
cri '6• ro ‘0 to
tal
hal
itel
(c91
0C0i
n)
BC
C
Me
•
tete
LS
O
CN
Af%
ICS
mat
o
se
ft,
'Met
al
1 1,
A.A
.0 1fl
11
111 I
e
nor
mop
s,
‘"\-
-&--
.
____
Fla
ns
Ovi
Vel
sW
C,
01
/at
end
ICS
HM
CO
(on.
)
Ico
.Ill,
./1.,-
"tcj
4\j
a "V
.'1
1/4‘,
1/4-
; % !.,, I1
tiI
,/1/
4.1U
k...
./1
-.N
"`
'P.
11c, NIN
....
tin ICC
CO
I
own
Ad
eall
Fig
ure
D.4
.Riv
erG
was
hat
Bel
mes
thor
pe.T
otal
habi
tata
ndflo
w.
Hab
itatt
ime
serie
sfor
brow
ntr
out
(adu
lt)
Val
ueso
utsi
dera
nge
ofm
odel
are
take
nas
the
valu
esat
the
limits
ofth
em
odel
WUA
(%
habitat
available)
0.5
0.2
0.15 10 20 30 50 70 80 90 95
% of time exceeded
Figure D5.I. This chart created by ignoringmissing values
50
20
10
5
2
1
1Adult Trout
1
I;
I
I
Summer Winter^ ...
50
20
5 10,-..as>5
..13n2—
f I<D0.5
0.2
0.1
•1
5 10 20 30 50 70 80 90 95% of time exceeded
Figure D5.2. This chart created by setting all1VUA values corresponding to flows outsiderange of the model to the WUAs at therespective limits ojthe model
1 ,...1 ...
II
I I I1
I IAdult Trout
Figure D5. River Gwash: duration curves for adult trout, examining effects of missing values
EA Technical Report W19 98
•
• •
•1,000
•500
•200
Flow
1
•
•
EL; 100
...
tc:`) 50
•
• 0
11-4 20
...
•
•
10
•
5
•
•
2
20 MI 511 , 0
90 95 • ce
66 67 68 69 70 71 72 73 74 75 76 77
Figure D6. River Ifye Yearkflow duration curves
EATechnicalReportW19 99
••••••••••••••••••••••••••••••••••
Adult brown trout
25
5 5 I0 20 30 50 70 80 90 05 % accedance66 67 68 69 70 7 I 72 73 74 75 76 77
. _
Figure 07. River Wye Yearlyhabitat duration curves
Fry brown trout (summer months only)
20
18
a 16Ct
14ct
co 12I I.\
10
6I0 70 30
Of, 67 OK mi 70 71
50 •0 KO7 7 71 74
99 ° exceedance7i 76 7' °
Figure 08. River Hit, }twilv habitat duration curves
EA Technical Report W19 100
á
••••••••••••••••
•
300
200
100
Fri50
30
20
66 67
5
68
1069
2070
Flow
3050
7172
7073
8074
90
75
..........
95 %exceedance
7677
••••••••••
FigureD9.RiverLarnbourn:yearlyflow durationcurves
•EA Technical Report WI9 101
••
wesseo
pire••••••••••••••••••••••••••
Adult brown troutrn
51020 890 9 o exceedance66 67 68 69 7(1 71 77 73 74 75 76 77
Figure 010. River Lambourn: yearly habitat duration curves
Fry / juvenile brown trout30
25
a)20
cct.—ct
cirs15
465
et
0 !..,:f 10
exceedance6h 6,4 69 70 7 7, 73 74 7 76 77
Figure 011. River Lumboum: yearly habitat duration curves
t Technical Report wig 102
-•
ww
ww
ww
ww
ww
ww
ww
ww
wW
WW
WW
WW
WW
WW
WW
WW
WW
••••-•••• ......
......
....
•• • . . .... ........
31,
Adult brown [Nut
r
Ju‘enlln Hov.m truttr
411
r
..........................
Figure 0I2. Ri er Wye. Flo \N spell anaksis of 16 years of daily flov, habitat data
t•A Fechnical Report W19 103
••••••••••••••••••••••••••••••••••
a
Flow700 —ii.
7600
>-.-2 0500.2
-orf, 400
......•
=. 200
; 300gc
..........gc
100,.
0
j_jj 600..)
".:77,500g
400iii-g
IR,
•:4300gf 200vc0
100\ itl 1.
•
•
•
•
•
•
•
••
700
600
-ra500
T400
‘;' 3005V00
100
00
10
10
20040I isnuI410). outman)) 4mallon, I' Al kVA shAted 41makoll I
Fryjuvenile browntrout
20 1040
Cumulaljconlin0ous dimanoso ("4011441 MudnM 4111411on
5(1
co.
Figure DI3. River Larnbourn: flow spell analysis of 16 yearsof daily flow ihabitat data
EA TechnicalReportW19 104
•
•
0 10 20 40 50 60
( .:onlinuous durations 0)40lImal sluIS duration)
Adult browntrout700
••••••••••••••••••••••••••••••••fr.•