Running Waters

Water flow at all scales

Sand-Jensen, K.

Published in: Running Waters

Publication date: 2006

Document version Publisher's PDF, also known as Version of record

Citation for published version (APA): Sand-Jensen, K. (2006). Water flow at all scales. In K. Sand-Jensen, N. Friberg, & J. Murphy (Eds.), Running Waters: Historical development and restoration of lowland Danish streams (pp. 55-66). Aarhus Universitetsforlag.

Download date: 07. Oct. 2021 Running Waters

EDITORS Kaj Sand-Jensen Nikolai Friberg John Murphy Biographies for Running Waters

Kaj Sand-Jensen (born 1950) is professor in stream ecology at the University of Copenhagen and former professor in plant ecology and physiology at the University of Århus. He studies resource acquisition, photosynthesis, growth and grazing losses of phytoplankton, benthic algae and rooted plants in streams, lakes and coastal waters and the role of phototrophs in ecosystem processes. Also, he works with speciﬁ c physiological processes, species adaptations and broad-scale patterns of biodiversity and metabolism in different aquatic ecosystems.

Nikolai Friberg (born 1963) is senior scientist in stream ecology at the National Environmental Research Institute, Department of Freshwater Ecology in Silkeborg. He has a PhD from University of Copenhagen on the biological structure of forest streams and the effects of afforestation. His main focus is on macroinvertebrates: their interactions with other biological groups, importance of habitat attributes and impacts of various human pressures such as hydromorphological alterations, pesticides and climate change. Also, he is involved in the assessment of stream quality using biological indicators and the national Danish monitoring programme.

John Murphy (born 1972) is research scientist at the Centre for Ecology and Hydrology, River Communities Group in the United Kingdom. His PhD from the University College of Cork, Ireland focused on stream macroinvertebrates associated with detritus in catchments of contrasting land use. His present research is on understanding the natural factors and human pressures inﬂ uencing macroinvertebrate communities in streams and rivers throughout the UK. Also, he works on predicting the consequences of future climate change and radionuclide contamination for macroinvertebrate assemblages in freshwaters. EDITORS: Kaj Sand-Jensen, Nikolai Friberg & John Murphy

Running Waters Historical development and restoration of lowland Danish streams

National Environmental Research Institute Ministry of the Environment . Denmark Running Waters Historical development and restoration of lowland Danish streams by Kaj Sand-Jensen, Nikolai Friberg & John Murphy (editors)

Layout and drawings: Juana Jacobsen and Kathe Møgelvang, NERI Graphics Group Translation: Anne-Dorthe Villumsen

Printed by: Schultz Graﬁ sk

ISBN 978-87-7772-929-4 Photografers:

Price: EURO 35 (incl. 25% VAT, excl. postage) Mette Dahl (43), Kaare Ebbert (103), Nikolai Friberg (52R, 84L, 84R, 88, 90, 144), Fyns Amt (26), Torben Geertz (143), Hans Ole Hansen (98, 124, 127T, 127B, 129B), Henrik Schmidt (33), J.W. Luftfoto og Sønderjyllands Amt (122), Dean Jacobsen (22, 52L, 104, For sale at: 106, 111), Per Søby Jensen (127M), Jørn Kjems (44), Anker R. Laubel (29, 42), Bent Ministry of the Environment Lauge Madsen (14, 16, 19, 36, 77, 82, 130, 132, 134), Jørgen Skole Mikkelsen (102), Alexander M. Milner (12), Bjarne Moeslund (129T), Christian Nielsen (97), Ole Fogh Frontlinien Nielsen (74), Finn E. Petersen (146), Kaj Sand-Jensen (51, 62, 64), Finn Sivebæk (87, 92, Rentemestervej 8 94, 96, 100, 142), Jens Skriver (cover, 54, 66, 81, 114, 117, 118, 120, 137L, 137R, 138), DK-2400 Copenhagen NV Martin Søndergaard (50), Ole Vestergaard (60, 61, 68T, 68B, 69, 71, 72, 107, 113, 140), Peter Wiberg-Larsen (76, 78, 141), Jan Kofoed Winter (24). Tel: +45 7012 0211 [email protected] Foreword

Many people are fascinated by the dynamic nature and heterogeneity of streams. Nothing is exactly the same from one second to the next and from one stream site to another. For scientists, however, the dynamic nature of streams is a constant challenge. To study streams and provide an accurate and reproducible description of the physical and chemical environment, and the biological life it supports, is a diffi cult, though rewarding, task. Among all the variability there are regular patterns and constraints which are possible to determine and use in practical management. Stream fl ow is a permanent downstream movement – turbulent and unpredictable. A small quiet and slowly fl owing stream can rapidly turn into a roaring river eroding the banks and transporting many tons of sediment following a few days of heavy rain. The stream can change direction and location if it is allowed to do so or has uncontrollable forces. Streams have this shifting gentle and angry face which we appreciate but also a temper we want to be able to predict to prevent unpleasant surprises. According to ancient Nordic mythology, the troll found a home in the cold water of streams when the glaciers melted under him 13,000 years ago. Ever since, he has remained the spirit of the stream who from time to time became furi- ous and takes a toll from the carriages, horses and drivers passing a stream. The gods, the prophets and their disciples who later occupied the country have been unable to drive him out of his hide in the pool under the bridge. Only here, out of sight and out of reach, is he safe, though doomed to stay in the wet element. We have made a great effort to describe and display important physical processes and biological creatures along and within the streams. Many people have helped us with photos. Fortunately they have never caught the image of the troll, which could have raised an omnivorous anger among his many hidden cousins distributed in the streams and rivers over the entire Earth. In the spirit of the Enlightenment this has become a book on the visible, predictable and comprehensible nature of Danish streams which can help us attain understanding and sensible management. The National Environmental Research Institute, Denmark, and director of research Kurt Nielsen have supported us with funds for printing and time for exceptional technical assistance from Juana Jacobsen and Kathe Møgelvang. We are deeply indebted to photographers, the research institution and the three persons above. We would also like to thank Jon Bass, John Davy-Bowker, Mike Furse, Iwan Jones, Martin Neale and Mattie O’Hare of the UK Natural Environment Research Council Centre for Ecology and Hydrology for their help in reviewing and improving the book.

Hillerød, Silkeborg and Dorchester, April 2006

Kaj Sand-Jensen Nikolai Friberg John Murphy Introduction 9

Lowland river systems – processes, form and function 13 1 Morten Lauge Pedersen Brian Kronvang Kaj Sand-Jensen Carl Chr. Hoffmann

Hydrology, sediment transport and water chemistry 27 2 Brian Kronvang Ruth Grant Carl Chr. Hoffmann Niels B. Ovesen Morten L. Pedersen

From spring to river – patterns and mechanisms 45 3 Kaj Sand-Jensen Tenna Riis Ole Vestergaard Bjarne Moeslund

Water ﬂ ow at all scales 55 4 Kaj Sand-Jensen

Aquatic plants 67 5 Tom Vindbæk Madsen Kaj Sand-Jensen

The terrestrial life of stream-dwelling insects 75 6 Peter Wiberg-Larsen

Macroinvertebrates and biotic interactions 83 7 Nikolai Friberg Stream ﬁ sh and desirable ﬁ sh stocks 93 8 Christian Dieperink Kaj Sand-Jensen

Water plants past and present 105 9 Tenna Riis Kaj Sand-Jensen

Improvements ahead for macroinvertebrates ? 115 10 Jens Skriver Hans Thiil Nielsen

A new development: Stream restoration 123 11 Hans Ole Hansen Annette Baattrup-Pedersen

Environmental state and research 133 12 Torben Moth Iversen

Streams and their future inhabitants 139 13 Kaj Sand-Jensen Nikolai Friberg

References 148

Index 156 8 RUNNING WATERS

NORWAY

Skagerrak

Ålborg Kattegat

River Gudenå River Storå Århus Herning North Sea

River Skjern Horsens

Vejle Copenhagen SWEDEN Jutland Esbjerg Kolding Great Belt Odense Zealand Funen Bornholm

Baltic Sea

Falster Lolland GERMANY

0 50 100 km Introduction 9

Introduction This book was in its original Danish version written for Danes with a general interest in nature or a speciﬁ c interest in streams and their biology. Whatever background, however, they were all Danes and as such familiar with our small country. Now we want to reach an audience outside our borders and have therefore made this introduction, which gives a brief background to Denmark beyond streams, macrophytes and macroinvertebrates.

Running waters focuses on geomor- the typology and public administration phology, hydrology and ecology of of Danish water courses. streams and rivers in Denmark and all Denmark is a small, lowland coun- anthropogenic impacts acting upon try (43,000 km2) with more than 403 them. The book was initially published islands. Nowhere in Denmark does the for a Danish audience, but when we distance to the coast exceed 50 km and showed it to colleagues in England, the coast line is very long (> 7,300 km). New Zealand, Estonia and Lithuania The peninsula, Jutland is connected they asked for an English version. to the European continent, while the They were impressed by the quality island of Funen and the large island of of illustrations and layout and wished Zealand are interconnected via bridges to learn more about Danish lowlands; (Figure 1). Zealand is separated from knowlegde streams which would be Sweden by the narrow strait, Øresund, applicable to lowland streams in their connecting Kattegat and the Baltic Sea. own countries. Moreover, they wanted Southwest of Zealand are two other to know more about the Danish policy main islands, Falster and Lolland, on pollution control of industrial and while the island of Bornholm is located domestic sewage and the initiatives far to the Southeast in the Baltic Sea. Figure 1 taken to bring channelised streams Denmark is a densely populated Map of Denmark with in the farmland back to their original and intensely cultivated country, the main regions, meandering form. inhabited by about 5.3 million people largest cities and three To meet this interest, we have (approx. 125 persons/km2). The capitol, largest rivers. produced an English version and John Copenhagen, is located in northeastern Murphy has kindly helped us editing Zealand, on the Øresund, and about it. However, to fully appreciate the 2 million people live in the greater book and facilitate its reading, we will Copenhagen area. More than 80% brieﬂ y introduce you to the basic geol- of Danes live in urban areas and the ogy and geography of Denmark and other densely populated regions in the 10 RUNNING WATERS

Box 1 Danish terms for running waters

“Vandløb” is both a technical and are “å” to us because they are not receives the most precipitation, while common Danish word for running suffi ciently big to become a “fl od”. Funen and Zealand hold one large waters of all kinds. To distinguish dif- You may notice that bæk is equivalent stream each. ferent sized watercourses in Danish to beck used in Yorkshire and Scotland Over very short distances in their we use the words “kilde” for spring, as a remnant of the infl uence of Scan- upstream parts, a few Danish streams “kildebæk” for small brooks, “bæk” dinavian settlers in the Viking period. may have relatively steep slopes for brooks, “å” for streams and rivers Å is used throughout the Nordic coun- (> 100 m/km), but true mountain and “fl od” for very large rivers. The tries, i.e. Sweden, Norway, Iceland and streams do not exist as there are no Thames and the Rhine would be the Farao Islands as the common name mountains in Denmark. The highest “fl od” to Danes, while the two largest for large streams. point is only 173 m above sea level. The rivers, River Skjern and River Gudenå majority of streams, therefore, have very shallow slopes (< 5 m/km). Denmark’s surface geology country are mid-Funen around Odense fl ooding. Increased sand erosion in consists of soft sediments of clay, sand and eastern Jutland around Kolding, coastal areas blocked streams and and gravel left by the two latest glacia- Frederica, Vejle, Horsens and Århus. further exacerbated the impacts of tions. During the second last glaciation, The western and northern parts of increased water levels. To counteract the entire country was covered by ice. Jutland have fewer people. the effects of increased water levels During the last glaciation, ending some Denmark is very suitable for and secure agricultural production, 13,000 years ago, the ice sheet glaciers agriculture as the climate and soils meadows were drained and streams covered all the islands and the north- are optimal for cultivating crops. The channelised and culverted. The agri- ern and eastern parts of Jutland. The climate is temperate Atlantic with fairly cultural area in Denmark peaked in the glaciers stopped in a prominent front, mild winters (0 ºC), not too hot sum- 1930s at 76%. whose two parts ran west-east from the mers (17 ºC) and an average annual Agricultural production is very present North Sea coast to mid-Jutland precipitation of 720 mm approximately effi cient with dense stocks of pigs and and north-south from mid-Jutland to evenly distributed over the year. Only dairy cows and intensive use of fertilis- the current German border (Figure 1). the island of Bornholm consists of ers and pesticides. In 2005 there were Only the south-western part of Jutland bedrock in some parts, while all other about 13 million pigs in Denmark with was left un-glaciated. Large rivers parts of Denmark have sandy and an annual production of about twice ran from the glacier front towards the loamy moraine soils. Therefore, almost that number. The main fertiliser is live- south-west between relatively shallow every part of Denmark has been culti- stock manure, and the main nutrient sandy hills left over from the former vated. Today the dominant land-use is sources of nitrogen (> 80%) and phos- glaciation. The glacial rivers left gravel agriculture (63%), while forests (mainly phorus (approx. 50%) to inland and and sand in West-Jutland while clay plantations) cover about 12% and other coastal waters derive from agriculture. was deposited further to the west in nature areas cover about 10%. Lakes Intensive agriculture and livestock pro- the central part of the present North and streams cover only about 1% of the duction, mostly pigs, are dominating in Sea. country. western part of Denmark i.e. in Jutland. The glacial history, therefore, There is a long history of landscape All Danish water courses are small explains the topography and surface alteration in Denmark. Deforestation and short. Only River Skjern, River geology of Denmark, and to some began early; by about 2,000 years ago Gudenå and River Storå in Jutland are extent the distribution of animals and approximately 50% of forest cover was suffi ciently large to deserve the predi- plants. West-Jutland is dominated lost. At the beginning of the 19th cen- cate river (Box 1). All other Danish by well-leached and carbonate-poor tury the area covered by forest reached water courses are small streams, brooks deposits of sand and gravel. East-Jut- a minimum of less than 4% and the and springs. Of the ten largest Danish land is dominated by poorly-leached substantial deforestation resulted in streams, eight are located in Jutland, and more carbonate-rich moraine increased groundwater levels and which has the largest surface areas and deposits of clay, sand and gravel. The Introduction 11

islands of Funen, Zealand, Lolland and features and hydromorphological ele- Falster are dominated by less-leached ments. From 2007 the counties will be and carbonate-rich deposits of pre- closed as part of a large-scale struc- dominantly clay. tural reform and regional, as well as Denmark is known for implement- national, monitoring activities will be ing the fi rst Ministry of Environmental transferred to the state. Protection (1971). Relatively intense To summarise: Danish landscapes ecological monitoring of streams also are intensely utilised almost every- started around 1970 and was intensi- where and the historical and confi ed even further around 1990. Today, temporary agricultural infl uence on it is the responsibility of the 14 Danish the morphology, management and counties covering all Denmark to set ecological quality of streams is very quality objectives for streams and strong. Most streams are small, but rivers. Each County Council politically more prominent streams and rivers decides quality objectives, which range exist in Jutland. Their biodiversity and from stringent objectives in streams of abundance of fl owering plants, mac- specifi c scientifi c interest to streams roinvertebrates and fi sh, mainly brown with eased objectives i.e. heavily trout, can be very high making Danish impacted streams. It is also the counties streams very important for sport fi sh- responsibility to monitor the quality of ing, recreation and overall biological streams in their region to ensure that quality among European lowland they fulfi ll the objectives set. Monitor- streams. ing is based on macroinvertebrates and the regional network comprises in total more than 10,000 sites. In addition to regional monitoring activities, a national monitoring programme for the aquatic environment, including streams, was launched in 1988 with revisions in 1992, 1997 and 2003. The programme is jointly undertaken by the Danish EPA, the National Envi- ronmental Research Institute (NERI), the Geological Survey of Denmark and Greenland (GEUS), the National Forest and Nature Agency and the 14 Danish counties. The counties undertake stream sampling but data storage, analysis and reporting are the responsibility of NERI. Today, the programme encompasses more than 800 monitoring stations covering all stream types in Denmark and monitoring includes three biological quality elements (macrophytes, macroinvertebrates and fi sh) as well as physico-chemical 12 RUNNING WATERS

Giant ice sheets and melt water have formed the landscape and river valleys. 13

1 Lowland river systems – processes, form and function Present day river valleys and rivers are not as dynamic and variable as they used to be. We will here describe the development and characteristics of rivers and their valleys and explain the background to the physical changes in river networks and channel forms from spring to the sea. We seek to answer two fundamental questions: How has anthropogenic disturbance of rivers changed the fundamental form and physical processes in river valleys? Can we use our understanding of ﬂ uvial patterns to restore the dynamic nature of channelised rivers and drained ﬂ oodplains in river valleys?

Danish rivers, their fl oodplains and At the glacier front in mid-Jutland, river valleys the melt water concentrated in several – formation and characteristics large rivers that fl ooded the moraine During the last Ice Age the forces of landscape in western parts of Jutland. fl owing water and erosion by glaciers The glacial melt water created large created the river valleys. The Danish washout plains where sand, gravel islands and eastern and northern parts and stones were deposited. Today of Jutland were covered by a gigantic these washout plains are heath plains ice cap during this Ice Age. Underneath acting as wide river valleys between the ice, water was fl owing in melt water the old moraine hills. The heath plains rivers from the base of the ice cap to are river valleys for some of the large the glacier fronts. Melt water fl owed contemporary rivers in Denmark. The in the valleys of the glaciated land- river with the greatest discharge in scape, thereby enhancing the landscape Denmark, the River Skjern, runs west features (i.e. hills and valleys) already from Nørre-Snede between Skovbjerg present. On its way to the glacier front, and Varde moraine hills and enters the fast-fl owing melt water eroded deep the sea in a large delta in Ringkjøbing valleys, while the slower- fl owing waters Fjord. deposited sand, gravel and stones as Since the last Ice Age rivers and ridges (hills). The outcome can be seen streams have eroded the pristine in the Danish landscape today, for moraine landscape and have formed example in the deep valleys of eastern fl oodplains and valleys along the Jutland where the River Gudenå and its rivers and streams (Figure 1.1). On the tributaries fl ow. In addition, the melting moraine hills exposed during the last of large ice blocks, left in the valley fol- Ice Age in western Jutland, the streams lowing the retreat of the ice cap, created have had approximately 100,000 years lakes such as those in the River Gudenå to interact with the surrounding land- valley. scape. This long period explains why 14 RUNNING WATERS

200

150 the valleys and fl oodplains are wider in quence of postglacial landmass move- western Jutland than in the relatively ment, land subsidence in the southern 100 young moraine landscape of eastern part of Jutland has created special con- (m) Denmark. ditions where the present-day rivers 50 In northern Jutland the post-glacial are fl owing in marine deposits and River valley width rise in sea level (Yoldia Sea) created continuously adjust their morphology 0 marine deposits near the shores. These to the rising sea level. On the island of 1234 have subsequently been exposed by Bornholm, the presence of bedrock cre- Stream order the tectonic upheaval of the landmass. ates an in-stream environment different They form the present-day wide river from anywhere else in Denmark. Figure 1.1 The river valleys increase in valleys, bordered by former sea cliffs width from the source to the sea – just as that now stand out as steep eroded The river valley – intimate inter- rivers do. Here is an example from the Brede hillsides, in this part of the country. actions between river and fl oodplain Stream catchment in southern Jutland. A good example of this is the Skals Rivers are naturally in a dynamic Stream, north of Viborg. As a conse- equilibrium with their valley and fl oodplain. The stream slope and fl owing water create the energy needed A journey along the River Skjern starts for transporting eroded soil particles at the brook (Svinebæk Brook at Lake from stream banks and hillsides in the Rørbæk) which later grows to a large catchment to the sea. Erosion is most stream (the River Skjern at Thyregod) prevalent in the upper river system, and ends at the mouth of the river in where many small streams are in close Ringkjøbing Fjord. contact with the surrounding landscape. Groundwater and drainage water from the catchment also supply vast quantities of dissolved substances to the streams. These are also subsequently transported to the sea. In the lower part of natural, unmodifi ed river systems, fl ooding of the river valley occurs regularly at high-fl ow events. During fl ooding vast quantities of sediment and organic debris are deposited on the fl oodplain. Close to the stream bank most of the sand is deposited as levees. Finer sediment and organic particles are deposited further away from the stream. The fl ooding waters, sediment and organic material are an important renewing source of nutrients to the fl oodplain. In natural systems these supplies of nutrients and sediment are essential for the growth and development of natural fl oodplain vegetation. In this way nature has created a buffer 1 Lowland river systems – processes, form and function 15

50 system delaying the loss of nutrients leaves meanders cut off and isolated 40 and sediment to the sea and enhancing on the ﬂ oodplain as small oxbow lakes. 30 the recycling of nutrients within the Vegetation and deposited material

(tonnes) 20 catchment. A substantial proportion of from the stream will eventually fi ll the the suspended material being carried lakes and create small wetlands on 10 by fl oodwaters can be deposited on the the fl oodplain. Oxbow lakes have a 0 fl oodplain and hence retained within characteristic mixed vegetation assem-

Sediment deposition/transport Dep. Transport Dep. Transport the catchment (Figure 1.2). blage, different from that on the rest of 23 Nov – 2 Dec 11 Jan – 20 Jan Streams in undisturbed lowland the fl oodplain, and are quite distinct 1992 1993 landscapes will naturally wind or features in our natural river valleys. Figure 1.2 Floodplain sediment deposition meander through the river valley. A The structure and composition of during high fl ow events can reduce the stream moves from side to side within fl oodplain soils are highly variable. The sediment transport to the sea. During two the river valley by eroding sediment on Brede Stream valley in southern Jutland 9-day fl ooding events in the lower part of the outside of meanders and deposit- has alternating layers of sandy stream the Gjern Stream (catchment area: 110 km2) ing the sediment on the inside of mean- deposits, peat and organic material on between 6% and 11% of the transported ders further downstream. It is a slow its valley fl oor (Figure 1.3). Deposits sediment was retained on the fl oodplain [1]. physical process, which evolves over of organic origin generally dominate decades or centuries. In some cases the soils in Danish river valleys. Decom- stream erodes at the base of the valley posed coarse plant material makes up sides, thereby slowly widening the peat layers and fi ne particulate organic fl oodplain. The constant lateral move- matter and diatom shells deposited in ment of rivers and streams occasionally shallow waters make up the gytja layers.

Figure 1.3 The Restored Channelised diversity of soil stretch 1994 stretch types in the river Southern side Northern side valley of the Brede 10 Stream is considera- 9 Ground surface 8 1954 ble. The channel of Peat soil Fine-grained sand Stream sediment Medium-grained sand 7 this 4-km reach was Gley soil Coarse-grained sand 6 re-meandered in Ferreous soil Clayey sand 5 1994 in an initiative Filling Peat 4 by the County of Sand drift Organic silt Ground level (m.a.s.l.) 3 Lake Filling 0 25 50 75 100 125 150 175 200 225 250 Southern Jutland. The river valley was drained and the river was channe- New meandered stream (1994) Undisturbed stream before 1954 lised in 1954. Oxida- Channelised stream (1954–1994) tion of the organic soil layer has caused a 0.5 m subsidence N of the ﬂ oodplain over the period of 100 m 40 years [2]. 16 RUNNING WATERS

During high fl ow events the natural unregulated watercourse fl oods the river valley and sediment and nutrients are deposited on the fl ood plain.

In total, 6,700 km2 (15%) of Denmark’s landscape on stream morphology Box 1.1 Stream order area consist of such poorly drained soils depends on soil type, topography, The systematic pattern of longitu- rich in organic layers. These soils are terrestrial vegetation and land use. dinally connected reaches increas- primarily found in the river valleys and Streams therefore form a unique ing in size at confl uences can be on raised seabed deposits in northern ecosystem because of the linear con- presented as a hierarchical system Jutland and in the salt marsh areas in nections within the network, the unidi- determined by origin and down- southern Jutland. These soils have been rectional fl ow and the intimate contact stream connection – this is known in high demand because when drained with the land. as the Stream Order [3]. they provide excellent nutrient-rich soils Many brooks are formed by springs for agriculture. located near the base of steep slopes, Stream order according to size where water leaves the groundwater Streams – a network of reservoir. The confl uence of several unidirectional fl ow springs forms a spring brook, which is 1 1 River systems form a fi nely branched characterised by stable discharge and network of many small, headwater water temperature. Further down- 2 1 streams that meet to form fewer and stream the confl uence of spring brooks 2 larger streams that fi nally merge into forms a stream, usually 0.5–2 m wide. 3 2 one large river reaching the sea (Box Streams increase in size as they fl ow

1 1.1). From upstream brooks to the downstream becoming larger, 2–4 m 1 downstream river, water is continu- wide. Even further downstream at the ously received from the surrounding conﬂ uence of the larger streams, the land. The inﬂ uence of the catchment river is formed. 1 Lowland river systems – processes, form and function 17

Box 1.2 Danish watercourses: length and location

There is never a long distance to the We have many both naturally and Watercourse Total length sea from anywhere in Denmark. Rivers, artifi cially created watercourses Small watercourses 48,000 km in the strictest sense, are therefore not in Denmark. The vast majority of (Width: 0–2.5 m) st nd very common. Denmark has only two watercourses are of 1 and 2 order Medium-sized watercourses 14,500 km large rivers: the River Gudenå, which is – these are called brooks. Medium- (Width: 2.5–8.0 m) rd th the longest river and the River Skjern, sized watercourses (3 to 5 order) Large watercourses 1,500 km which has the highest discharge (Box – the streams – are fewer in num- (Width: >8 m) 1.2). bers and we only have a few large Total 64,000 km watercourses. The longest Danish Flow and substrata from spring watercourse, River Gudenå, is 176 km to river long. Based on physical appearance With increasing distance from the and biotic communities only the River spring, the stream usually becomes Gudenå, River Skjern and River Storå wider, deeper and transports more can be classifi ed as large rivers. water (Figure 1.4 and 1.5). Large rivers often originate in mountains, run down The map shows the 20 largest watercourses in Denmark measured by catchment the mountain slopes and across plains area. towards the sea. Along the course, the slope decreases [4]. This is generally Uggerby Stream (18) also the case in the short river systems Rye Stream (11) in Denmark, where the steepest slopes are usually found in the springs and Skals Stream (10) brooks, and the lowest slopes in the rivers close to the outlet [5, 6]. How- Karup Stream (8) ever, on the local scale Danish streams Lindenborg Stream (19) follow a more irregular course with shifting reaches of steep or low slopes. This is primarily due to the presence of River Gudenå (1) lakes, which act as hydraulic thresh- Grenå Stream (15) olds and re-set the natural dynamics River Storå (3) of the fl owing streams. Because of the local irregularities reaches with steep Halleby Stream (13) slopes and high physical stress may River Skjern (2) occur anywhere along the course. Vejle Stream Gravity, acting on the slope of the Varde Stream (4) (20) water surface, drives the downstream Sneum Stream (14) fl ow. The mean current velocity along Konge Stream (17) reaches also depends on the resistance Ribe Stream (6) Brede Stream (12) to fl ow exerted by the streambed, banks and plant surfaces. Since upstream Vidå Stream (5) brooks are shallow and narrow, water fl ows in intimate contact with the Odense Stream (9) streambed and banks. So in spite of a Tude Stream (16) steeper slope, the mean current velocity Suså Stream (7) is usually lower in small Danish brooks than in large rivers [3, 4]. 18 RUNNING WATERS

Box 1.3 Lowland channel patterns

Figure 1.4 Idea- Classifi cation of channel patterns in lowland streams. The 100 lised graphs illustra- channel pattern is the result of many processes interact- ting how the large ing at many scales, and a classifi cation scheme provides an number of small overall insight into the dominant processes and param- 10 headwater streams eters (modifi ed from [5]). gradually merge No. of reaches 1 into fewer streams gravel+stones 12345 of higher order (top), and higher 10,000 discharge (bottom). 1,000

100 gravel (l/s) Slope fine sand, silt

Water flow Water 10

Bed material size 1 silt 12345

Stream order

Discharge/Catchment area

1.00 e 0.75 Local ﬂ ow conditions above the Channel patterns – fundamental d c streambed regulate erosion and sedi- principles of river morphology 0.50 (m)

Depth b mentation of particles. Although we Stream slope, water discharge, sedi- 0.25 a may intuitively assume that steep head- ment supply and grain size distribu- 0 waters have coarse gravel and stone tion of transported sediment, control 0–2 2–4 4–6 6–8 >8 substrata while large rivers have fi ne- the channel planform pattern in rivers grained sand and mud, this does not and streams. Since these parameters 0.4 apply to Danish lowland streams – nei- change systematically through the river d d 0.3 ther in their original unregulated state channel system, the same would be c nor in their contemporary regulated expected from the channel planform 0.2 b (m/s) a state. In a nationwide study of 60,000 pattern. In the following section we 0.1 plots (25 × 25 cm) in 350 reaches across examine how the link between channel Current velocity Current 0 75 streams, we did not observe syste- form and physical processes generates 0–2 2–4 4–6 6–8 >8 matic changes in the main substrata the distinct channel patterns. Stream width (m) along the stream courses. Sand was the Stream channel patterns have most common bottom substratum in traditionally been classifi ed as straight, Figure 1.5 There are systematic changes both small and large streams (approx. sinuous or meandering [3, 7, 8] (Box in mean depth and current velocity along 40% of all plots). In streams less than 2 1.3). The straight and sinuous channels the river continuum from source to outlet. m wide, gravel and stones were found are generally found in the upper parts Summer measurements in 350 Danish in 30% of the plots, while 23% of the of the lowland river systems where the stream reaches illustrate the overall pat- plots in streams more than 8 m wide channel slope is high and discharge terns. Different letters mark statistically had sediments dominated by gravel is low. These streams are often found signifi cant differences. and stones. Mud was also frequent in in the moraine landscape and the small (32%) and large streams (23%). stream bed substratum is dominated 1 Lowland river systems – processes, form and function 19

can transport a wider range of sediment sizes, resulting in signiﬁ cant sedi- 20 ment transport along the streambed. 10

This in turn leads to the development Loamy 5 moraine

of riffl es and pools characteristic of the (m) sinuous channel. Further downstream, Width with greater stream power, the char- Sandy wash out plain acteristic lowland meandering chan- 1 nel is formed. In natural sinuous and 1 5 1050 100 500 The channelisation of Danish streams meandering streams erosional zones 2 increased the slope of new straightened occur at regular intervals. Along the Sandy wash out plain stretches. Trapezoid weir-like constructions outside of the meander bends sediment 1 were built in many streams, to dissipate is eroded and deposition occurs further 0.5 energy from the increased slope over short downstream on the inside of meanders. (m) segments of the stream. Unfortunately Various sediment sizes can be trans- Depth Loamy moraine these constructions also act as obstructi- ported, depending on the strength of ons limiting the free migration of fi sh and stream power. Fine sediments such as 0.1 macroinvertebrates. Today many of these clay, silt and fi ne sand are transported 1 5 1050 100 500 obstacles have been transformed into riffl es in the water column as suspended Catchment area (km2) by local water authorities. solids, whereas coarse sand and gravel are transported along the stream bed. Figure 1.6 Natural channel geometry as a Coarse sediment is concentrated in the function of catchment area varies between equally by sand and gravel/stones riffl es, where fl ow divergence cause different landscape types in Denmark [9]. with scattered fi ne sediments. Ero- stream power to decrease locally, leav- The smaller streams on the sandy wash sion dominates in this part of the river ing the fl ow incapable of moving coarse out plain in western and southern Jutland system and material is continuously particles such as gravel and stones. Fine are generally deeper and less wide than added to the stream from bank erosion sediment can be captured between the the streams on the loamy moraine soils in and overland fl ow. The combination of large particles forming a very com- eastern Jutland. For streams in catchments sandy moraines and addition of mate- pact riffl e structure typically found in larger than approx. 100 km2 the relations- rial ensures also a continuously high lowland streams [8]. The fi ne sediment hips are reversed. proportion of sand on the stream bed. is deposited on the inside of meanders Stream power is insuffi cient to remove where current velocity decreases – the the coarse material, and the stones and depositional areas on the inside of and variations in climatic conditions gravel are left in channel, whereas clay, meanders are known as point bars. [1]. Since geology affects the ability silt and sand are transported further of the soil to be eroded, and climate downstream as suspended solids in the The cross section and groundwater conditions affect water. Within the streams, sidebars of Channel pattern and cross sectional the amount of water drained through coarse material are formed, with areas profi le change as you travel down the channel, the hydraulic geometry of fi ne-grained sediment deposited in through the river system. With increas- relationships can be expected to vary between. In streams with very steep ing distance from the source, depth, signifi cantly among streams in differ- slopes the streambed is made up of a width and current velocity vary as a ent parts of Denmark. series of steps dominated by gravel function of discharge (Box 1.4). These On the sandy washout plains of and stones and fi ner sediments in relationships are known as the hydrau- western Jutland, streams receive a large pools between the steps. As discharge lic geometry of the stream and vary proportion of groundwater all year increases, stream power also increases depending on geological and geomor- round. These streams tend to be deeper despite the lower slope, and the stream phological conditions in the catchment and less varied in their depth than 20 RUNNING WATERS

Box 1.4 Hydraulic geometry

Systematic variations in the physical proper- 100 In-stream habitats ties of the channel and ﬂ ow through the – predictable and variable 10 river system can be described using a set of Running water actively forms the physi-

equations, known as the hydraulic geometry (m) 1 cal habitats within streams. We have relationships [3].

Stream width Stream already shown how substratum com- 0.1 position varies through the river system w = aQb d = cQf U = kQm 0.1 1 10 100 from source to sea. Now we take a closer w = aAb d = cAf U = kAm look at the variations at a ﬁ ner scale, 100 within a given stretch of stream. Where: w is the bankfull width, d is the bankfull In all lowland stream types, from 10 depth and U is the mean current velocity at straight to meandering, the stream

bankfull discharge, Q is the bankfull discharge, (m) 1 habitats alternate between sections A is catchment area, a, c, k are constants and of high and low current velocity. In

Stream depth Stream b, f, m are exponents. Note that for any given 0.1 areas of high current velocity, coarse relationship b + f + m = 1 and a × c × k = 1. 0.1 1 10 100 sediment dominates, whereas ﬁ ne sediment dominates in areas of low The relationships are indicated as log-log plots 100 current velocity. The longitudinal of the dependent variable width, depth or distance between successive areas of 10 mean current velocity as a function of the dis- high current velocity is approximately

charge (or catchment area). The constants and (m/s) 1 5–7 times the width of the stream. exponents reﬂ ect properties of the catchment The ultimate development of this from where the stream water is drained. Differ- 0.1 habitat variation is the rifﬂ e and pool

Mean current velocity Mean current ent river systems in different parts of the world 0.1 1 10 100 sequence, which is a dominant feature will thus have different constants and expo- Bankfull discharge (m3/s) of all meandering streams (Box 1.5) nents refl ecting differences in climate, geology Riffl es and pools are distinctly dif- and geomorphology. ferent habitats with respect to substratum, depth and current velocity (Figure Exponents Reference b f m 1.7). These fundamental physical differences are also refl ected in the Upland river (Appalachians, USA) 0.55 0.36 0.09 [9] species composition and abundance of Great-plains river (Mid-western USA) 0.50 0.40 0.10 [3] macroinvertebrate communities in the Lowland river Denmark ( River Skjern) 0.48 0.44 0.08 [10] two habitats [12]. At an even fi ner scale, there are signifi cant variations in substratum, depth streams in the moraine landscape, as schemes. The hydraulic basis for the and current velocity within the distinct the catchment area increases. Further- functioning of different river ecosys- habitats and the varied nature of more, moraine streams are generally tems can be compared, using constants small-scale habitat features underlines wider for any given catchment area and coeffi cients from hydraulic geome- the heterogeneous nature of lowland (Figure 1.6). These generalities persist try relationships. These provide a valu- stream (Box 1.6). However, local habitat up to a certain catchment size, after able tool for assessing the variations structure within any stream reach still which the relationships between catch- in overall physical habitat conditions depends in part on large-scale condi- ment area and depth and width are at the river system or catchment scale. tions such as fl ow regime, and local shifted. However, if we want to know anything conditions such as river valley slope, Such general relationships can be about the actual habitat conditions meandering and riparian vegetation used to design channels with natural within the stream we have to look at structure. Variations in these control- dimensions in Danish river restoration fi ner scales (Box 1.5). ling factors potentially infl uence the 1 Lowland river systems – processes, form and function 21

Box 1.5 Channel morphology

Pool Riffle Current velocity, depth and streambed channels in the upper parts of the 100 substratum vary in a predictable river systems have sinuosities between 80 manner in natural stream channels. 1.05 and 1.5 and naturally meander-

60 The distance between two neighbouring channels have sinuosities higher ing rifﬂ es in a meandering channel is than 1.5. The sinuosity of the streams (%) 40 5–7 times the channel width. Straight in Denmark has been used to quantify 20 channels exhibit identical alternating the total length of natural channels

Substratum coverage 0 patterns. in all river systems [11]. The total Coarse gravel The channel sinuosity is deﬁ ned estimated length of channels in Den- Fine gravel as the length between two points mark was found to be approximately Sand along the deepest part of the channel 2,000 km. Almost half of these natural (known as the thalweg) divided by the streams are found in the western and 100 distance along a straight line between southern part of Denmark. 80 the points. Natural straight or sinuous

(%) 40

Percentage Percentage 20

0 Mud Pool 0–20 20–40 40–60 >60 Deposition Deposition Water depth (cm) Gravel Gravel Riffle 100 Gravel+sand Meander length 80 Sand/gravel Width 60 Depth Erosion Erosion (%) Gravel 40 Sidebar Meander length

Percentage Percentage 20 Mud 0 0–20 20–40 40–60 >60

Current velocity (cm/s)

Pool Riffle

Figure 1.7 Substratum, depth and current velocity of riffl es and pools in a lowland habitat structure between any riffl e in a landscape conditions infl uence the gen- stream in Denmark. Pools are deeper areas stream and cause signifi cant differences eral pattern of channel sinuosity. At the in the streams with low current velocity and in the micro-habitat structure both channel reach-scale, riffl es and pools fi ne substratum. In contrast riffl es are the within and among riffl es. These small- alternate, creating different environ- fast velocity and shallow areas dominated scale differences in habitats clearly ments. Within these distinct habitats by gravel substratum. affect the composition and abundance there is considerable fi ne-scale vari- of the local biotic communities [13]. ability in habitat conditions. And yet At fi rst glance lowland streams are despite this small-scale heterogeneity, physically homogenous with consist- we are capable of predicting general ent and predictable variations in their patterns in habitat structure and biotic physical features from source to sea. communities throughout the river Local hydrological, geological and system from source to sea. 22 RUNNING WATERS

Box 1.6 Small-scale variability in rifﬂ es

Variations in substratum, depth and environment that characterises in- Human impacts on Danish rivers, current velocity in riffl es are con- stream habitats in streams around fl oodplains and valleys siderable in lowland streams when the world. Large-scale characteristics River valleys and streams cut through surveyed in detail, as here from the of the individual habitats (riffl e/ pool) the countryside as corridors connect- Tange Stream, Denmark. Small-scale usually hide considerable variations at ing the land and sea by supplying a variations are part of the dynamic smaller scales. green vein for transport of water, sediment, nutrients, plants and animals. Streams, river valleys and fl oodplains Upstream riffle have been extensively exploited since Depth (cm) Velocity(cm/s) Substratum Stone + the fi rst humans inhabited Denmark 45 70 coarse gravel 40 60 after the last Ice Age. In the beginning, 35 50 Fine gravel river valleys acted as ideal places for 30 40 25 30 Sand settlements because of the possibility of 20 20 15 10 combining fi shing and hunting. Later, 10 0 5 streams and rivers became valuable 0 sources of energy for water mills. Dams that were constructed on many watercourses now obstruct the free fl ow of water from source to sea. Larger streams and rivers were important means of transport, e.g. barge transport from Silkeborg to Randers on the River Gudenå during the 18th and 19th century, and are still widely used for Downstream riffle recreational canoeing today. Fish farms Depth Velocity Substratum were established in many river valleys during the 20th century and they used the stream as a source of water for breeding and rearing trout. However agriculture has caused the most radical changes to the rivers and fl oodplains. For centuries fl oodplains 1 m were used for cattle and horse graz- 1 m ing and also for haymaking, used for winter fodder. Over the past 100 years

Streams are highly variable over small scales. 1 Lowland river systems – processes, form and function 23

Box 1.7 Habitats in channelised and restored reaches agricultural practices have intensifi ed, In channelised rivers and streams uni- spawning grounds have been lost. primarily due to increased crop produc- form cross sections with steep banks The combination of these actions has tion. This has resulted in drainage and have replaced the natural irregular reduced the available habitats for construction of ditches in river valleys cross sections. The uniform physical macroinvertebrates and fi sh. and the straightening and channelisa- conditions have reduced in-stream tion of many streams, to increase drain- habitat diversity in many Danish The habitats are more varied in a age effi ciency. Advanced technology streams. The dredging activities have restored reach of the Gelså Stream increased the development of drainage removed coarse substratum (gravel compared to an upstream channelised schemes, in particular by means of tile and stones) from the stream and fi sh reach [14]. drainage throughout eastern parts of Denmark. Many small streams in the upper river systems were culverted Channelised Restored in order to ease the use of agricultural reach reach machinery. Physical changes to the 6.34 m 6.00 m river valleys have altered hydrological 0 m conditions and affected both vegetation 0 m and animals. As a consequence of the 5 m drainage schemes carried out during 5 m the last century, Denmark has now 10 m almost no pristine streams and river 10 m valleys. Channelisation and straighten- 15 m ing have physically modifi ed more than 15 m 90% of the stream network of 64,000 km 20 m (Box 1.7). 20 m

The extensive fl oodplain drainage 25 m has caused signifi cant subsidence of Flow direction 25 m river valleys soils. This phenomenon 30 m 30 m is caused by decomposition of the 40 m 35 m peat layers when exposed to atmos- Habitats 35 m pheric oxygen. Subsidence levels of Riffle 0.5 m are very common in many river Pool 40 m Run with sandy substratum valleys and levels of up to 1 m have Run with gravel substratum Edge been measured in the lower part of the Flow River Skjern system, even though this direction Measurements area was drained as late as the 1960s [15]. As a consequence of the subsidence, the beds of ditches and streams have had to be further lowered and tile disturbance to stream conditions. This time straightening of the streams has drainage has been renewed. maintenance is currently still required reduced the stream length, which leads Large-scale channelisation and in many streams. The recurring need to an increase in stream bed slope. In straightening of streams has increased for removal of sediment from the order to compensate for the increased sediment erosion and mobilisation. streambed has profound consequences slope weirs were constructed in the In order to maintain drainage capac- for stream fl ora and fauna. Streams streams, whereby most of the energy ity in channelised streams, excess have gradually become wider and was concentrated in a few large weirs. sediment had to be removed from deeper than they would have been These weirs have acted as obstructions the streambed, causing continuous under natural conditions. At the same to the free movement of animals and 24 RUNNING WATERS

plants. Over 140 such obstructions were found within a restricted area such as Sønderjylland County (3,940 km2) in the late 1980s. The repeated disturbances caused by dredging helps to increase sediment transport both in terms of suspended transport and transport along the streambed. Over a period of 3 months following maintenance work on the Gelså Stream, sediment transport increased by 370 tonnes of which 280 tonnes was registered as bed-transported material [16]. An increase of this magnitude can have devastating effects on the stream biota due to sand intrusion into salmonid spawning grounds and general habitat degradation for macroinvertebrates.

What initiatives are needed to reverse the physical degradation of rivers and fl oodplains? Public interest in restoring the natural dynamics of streams and fl oodplains has been growing over the past 15 years. The best example of a completed project is the restoration of meanders along the River Skjern and the re-estab- River Skjern is the lishment of wetlands on its fl oodplain. largest restoration Prior to these initiatives, vast sums of project until now. restoration of spawning grounds that were lost during the money had been invested in improv- period of continuous dredging of stream sediments. The ing water quality in Danish rivers effective lack of restoration of spawning grounds is made and streams. As water quality has very diffi cult by the defeatist attitude towards the excess improved it has become increasingly continuous transport of sand into the streams, which causes clear that it is now primarily the physi- rapid siltation of the gravel beds. It is obviously necessary cal degradation of the rivers and fl ood- to track down, identify and combat the excess sediment plains that limits biological diversity. delivery and thereby reduce the sediment transport. Actions Over the last 20 years hundreds of could encompass a more effective enforcement of the exist- projects have focused on removal of ing compulsory 2 m-wide uncultivated buffer strips along obstructions in streams. These projects streams or the establishment of wider buffer strips, with no have enhanced the opportunities for agricultural production, to prevent sediment reaching the salmonoid fi sh to reach their upstream channel. spawning areas. Unfortunately, these The only possible way to re-establish more natural physi- efforts have not been supported by cal and biological conditions in the thousands of kilometres 1 Lowland river systems – processes, form and function 25

of streams is to change maintenance procedures, including timing, frequency and intensity of dredging and weed cutting. Future maintenance must include gentle procedures in those parts of the river systems where the potential for the development of diverse fl oral and faunal communities exists. Over the past 20 years maintenance procedures in many streams administered by municipalities and counties, have been changed towards gentle and selective weed cutting that leaves some vegetation in the channel and on the banks, and dredging activities have been reduced. Docu- mentation detailing the procedures that have the most positive effect on biodiversity is, however, still lacking. There is an urgent need for experimental studies in different Danish stream types to determine which maintenance procedures are most successful at sustaining and improving stream biodiversity. In order to proceed from here, a holistic approach to river and fl oodplain restoration is required. We need to see the river valley, the fl oodplain and the stream as one unit in a landscape mosaic. Information on the physical diversity of the river valley and its streams therefore needs to be integrated in a system allowing classifi cation of rivers and river valleys according to geomorphology, hydrology and geological setting. The ultimate goal of our restoration efforts should be based on physical and biological data on the structure of the few remaining undisturbed streams in Denmark. The river and river valley classifi cation systems in combination with the established biological reference conditions of our streams in different parts of the country will be an important tool for future decisions on how and where to implement restoration schemes. We also need to gather both short-term and long-term information on how the river and fl oodplain ecosystems respond to the restoration. The future success of our restoration effort will rest on our ability to document ecosystem responses to past restoration efforts and subsequently to improve our methods based on this knowledge. 26 RUNNING WATERS

Intensive agriculture domintate in Denmark leaving very few meandering streams with wide corridors functioning as buffers for nutrients. 27

2 Hydrology, sediment transport and water chemistry The physical and chemical quality of Danish streams has been entirely altered by the dramatic changes in anthropogenic pressures occurring over the last 150 years. Large-scale drainage schemes for reclamation of agricultural land, organic wastes from an increasing urban population, soil and bank erosion giving rise to excess sediment transport, leaching of nutrients and organic micropollutants are among the suite of anthropogenic pressures affecting the streams. The last three decades have been spent on introducing several Governmental and Regional Action Plans for combating these pressures and improving stream ecology.

Pollution history of streams to surface water from larger point Less than forty years ago the main pol- sources (>30 person equivalents) lution problem in Danish streams was during the 1970s and early 1980s in the discharge of organic matter from most Danish River Basins (Figure 2.1). ineffi cient sewage treatment plants, Hydrological monitoring of Danish and from industries, freshwater fi sh streams was initiated already in 1917 farms and small settlements without where the fi rst gauging stations for treatment facilities. Another pollution daily recordings of water stage were problem was the spill of liquid manure installed, which, combined with from agricultural farms that could acci- bimonthly discharge measurements dentally kill all life in a watercourse. permits estimates of daily discharge. Increasing public awareness of Since then the hydrologic network has water pollution problems resulted in become denser and currently more the establishment of the Danish Minis- than 400 hydrological stations are in try of Environment – the fi rst ministry regular operation [1]. Monitoring of of its kind in the world. The Danish water chemistry in Danish streams regional authorities were selected as did not begin until the mid-1960s as responsible for surveying the chemical part of the fi rst Hydrological Decade. and biological quality of groundwater Monthly sampling was initiated in and surface water and they were com- four large rivers in different regions of mitted to establish regional plans for Denmark. During the 1970s and early water quality objectives and to provide 1980s, several stations for monitoring an overview of the ecological state of water quality were established by the all streams and lakes every four years. Danish regional authorities in order to Substantial investments in improved measure the state and trend in water water pollution control resulted in chemistry, and the main pollution decreased discharges of organic matter sources of nutrients and organic matter. 28 RUNNING WATERS

Major implemented measures to diffuse nutrient pollution

1985 NPO Action Plan • Elimination of direct discharges from farms • Livestock harmony at the farm level Streams 231 water 1987 Action Plan on the Aquatic Environment I chemistry stations • 9-month storage facility for slurry • 65% winter green ﬁ elds • Crop and fertiliser plans

1991 Plan for Sustainable Agricultural Development • Standard nitrogen fertilization values for crops • Standard values for nitrogen in animal manure • Required utilisation of nitrogen in animal manure (30-45%) • Fertiliser accounts at farm level

1998 Action Plan on the Aquatic Environment II • Demands for an overall 10% reduction of nitrogen application to crops • Demands for catch crops • Demands for transforming 16,000 hectares farmland into wetlands • Reiterated demands for utilisation of nitrogen in animal manure (40-55%)

2003 Action Plan on the Aquatic Environment III • Halving of the Danish P surplus before 2015 • Establishement of 50,000 ha buffer zones to reduce P losses from agricultural ﬁ elds • Reiterated demands for catch crops and utilisation of N in animal manure

Table 2.1 List of Governmental Action Plans implemented in Den- Figure 2.2 Map showing the Danish stream monitoring network mark to reduce nutrient losses to the aquatic environment. for measuring hydrology and water chemistry according to a regular sampling protocol.

Figure 2.1 Trends in the concentration of BOD in the largest Danish streams during the period 1978–2004. The distribution of water quality stream stations in Den- 10 mark were, however, very sparse and not well distributed

/l) geographically until 1987 when around 50 stations were in 2 8 regular operation.

(mg O The Danish Parliament has adopted several plans for 6 improving the quality of water bodies since the early 1980s (Table 2.1). Following the ﬁ rst Action Plan in 1987 a national 4 monitoring network was established covering all water bodies. The monitoring network was put into operation 2 in 1988. The plan was to regularly monitor hydrology and

Concentration of BOD water chemistry at more than 270 stream stations throughout 0 the country (Figure 2.2). The monitoring network enabled 1980 1985 1990 1995 2000 2005 the Danish regional and national authorities to establish a Suså Stream Odense Stream nationwide state of the hydrology and water chemistry of River Gudenå River Skjern Danish streams and assess the temporal trends and effects of the various Environmental Plans adopted [2]. 2 Hydrology, sediment transport and water chemistry 29

Box 2.1 Draining of agricultural land

Large amounts of ammonium, dissolved Before After reactive phosphorus, sulphate and iron draining draining are leached from former wetland areas Ammonium 0.06 114 when these are drained for agricultural (NH4-N) purposes. Leaching of substances is high- (kg N/ha) est just after draining and declines during Total iron – 536 the following years when mineralisation (Fe) (kg/ha) rates decline and the soil level shrinks. Sulphate 28 1,875

(SO4-S) The two examples show reclamation of (kg S/ha) agricultural land on the ﬂ oodplains of the Orthophosphate 0.04 0.20

large River Skjern, which was straight- (PO4-P) ened and channelised in the early 1960s. (kg P/ha)

Net leaching of ammonium, iron, sul- 12 phate and dissolved reactive phosphorus 10 from a newly tile-drained wetland area along the large River Skjern. 8

6 Following drainage work along the main (mg/l) river channel the concentration of total 4

and dissolved iron increased dramati- concentration Iron 2 cally in the artiﬁ cial, excavated drainage 0 Approxymately 50% of the agricultural land channel that received pumped water 1975 1980 1985 1990 1995 in Denmark is drained and on loamy soils from the reclaimed ﬂ oodplain rich in with tile drains. pyrite. The elevated concentrations of Total iron Dissolved iron iron decreased slowly during the post drained period.

The changing landscape rural areas triggered large scale drain- drainage activity created water quality The Danish landscape has undergone age of most of the Danish land area problems in streams due to enhanced major changes during the last 100–150 through major drainage schemes that mineralisation of organic matter releas- years. These changes dramatically led to straightening and channelisation ing excessive nitrogen and phosphorus altered the physical conditions, hydro- of more than 90% of all Danish streams to surface waters. Another major ecological regime and water chemistry (see chapter 1). Moreover, numerous logical problem in Danish streams was of the streams. Sewerage systems and 1st order streams were culverted to created by drainage in riparian areas sewage treatment plants were installed facilitate easier access to fi elds for farm with a high content of pyrite. Here in most Danish towns during the 20th managing. Detailed drainage schemes, soluble iron was formed in the former century but in the beginning only with including excavation of ditches in the anoxic but now oxic soil layers and mechanical treatment and the waste predominantly sandy riparian areas transported to streams giving severe water were discharged into the nearest and installation of tile drains in the problems due to acidifi cation and water body, typically streams, thereby predominantly loamy riparian areas, precipitation of iron oxy-hydroxides, creating excess organic pollution for broke the natural interaction between ochre (Box 2.1). Currently more than kilometres downstream. The intensi- groundwater and surface water in 400,000 ha of former wet riparian areas fi cation of agricultural production in riparian areas. Moreover, the intense have been artifi cially drained and are 30 RUNNING WATERS

in agricultural production. Agriculture mainly as arable land, now occupies more than 60% of the Danish land area (Figure 2.3).

Urban areas Agriculture The hydrological cycle Natural areas Although Denmark is a small country Forest Wetlands there is a large west-east gradient in Lakes precipitation from an annual precipitation of more than 1,200 mm in western Jutland to an annual precipitation of less than 600 mm on the island of Zealand. Figure 2.4 shows the water balance in Denmark. The precipitation gradient results in a decreasing average and minimum runoff in the streams from west to east in Denmark (Figure 2.5). The west-east gradient in the water balance also has a strong impact on the median minimum runoff in Danish streams which is very low on south-eastern Zealand and Bornholm, reducing the ability of streams to dilute discharges of pollutants from point sources, re-aerate the water and ensure the biological quality during summer. Streams in eastern Denmark are therefore very sensitive to drought years and many small 1st order streams (see box 1.1) are at risk of drying out. This Figure 2.3 Land use in Denmark is classiﬁ ed in 6 main classes. Agriculture is the main land use all over the country.

Jutland The Islands

Precipitation Water abstraction Precipitation Water abstraction 940 mm 15 mm 750 mm 15 mm Figure 2.4 Water balance for Jutland and Evaporation Evaporation the islands during the period 1971–2000. 525 mm 540 mm Streams in Jutland receive most water

Q Q and have the largest interaction between u u G i G ic ck Runoff r k Runoff ro f o fl groundwater and surface water. Streams on u lo u o m n w m 395 mm n w 215 mm d ( m d (d 0 m w drain) 45 w rain) 9 the islands are more dominated by water at at er 395 mm er 125 mm discharging from the unsaturated, often tile-drained zone. 2 Hydrology, sediment transport and water chemistry 31

phenomenon has locally been acceler- Figure 2.5 Maps ated by abstraction of groundwater for Mean runoff showing the (mm/year) (l/s · km2) water supply to large cities, especially 50 1.6 average annual around the capital of Copenhagen. 100 3.2 runoff and median 150 4.8 Danish streams have been catego- 200 6.3 minimum runoff 250 7.9 in Danish streams rised into 5 different hydrological 300 9.5 regime types based on differences in 350 11.1 during the period 400 12.7 hydrology (Table 2.2). The categories 450 14.3 1971–1998 [1]. include 2 small stream types and 3 500 15.9 550 17.4 larger stream types and the regime 600 19.0 650 20.6 types are clearly distributed accord- 700 22.2 ing to region. The stable hydrological regime is mostly found in the northwestern part of Jutland, the moderately stable regime in central and eastern Jutland and the unstable ﬂ ashy stream types on the islands of Funen, Zealand, etc. to the east [1]. The hydrometric stations having been in operation since 1920 in Danish streams show a general upward trend in average annual runoff during the last 100 years (Figure 2.6). This trend can be explained by a simultaneous upward trend in annual precipitation during Minimum runoff (l/s · km2) the same period. Changes in land use, 0 drainage of agricultural land and extrac- 0.2 0.5 tion of drinking water from groundwa- 1.0 1.5 ter have also inﬂ uenced the hydrology 2.5 of Danish streams. An example is Århus 4.0 6.0 Stream near the largest city in Jutland 8.0 where extraction of drinking water has 11.0 reduced annual average runoff during the last 100 years (Figure 2.7). Most climate scenarios predict a wetter and warmer climate in Den- mark for the next 50–100 years [4]. Such a climate change will alter the water balance and the runoff in Danish streams. An estimate of the importance of climate change has been calculated by using a precipitation-runoff model, a modelled downscaled (25 km grid) 30-year time series of climate data for a reference period (1961–1990), and modelled data for the period 2071–2100 32 RUNNING WATERS

Small streams Large streams Stable Unstable Stable Moderate stable Unstable hydrological hydrological hydrological hydrological hydrological regime regime regime regime regime

Region Jutland The Islands Northwestern Central and The Islands Jutland southern Jutland Mean catchment area (km2) 82 59 834 394 472 Mean discharge (l/s) 717 406 11,300 5,400 4,100 Average frequency of daily discharge being 1.4 4.9 0 0.2 1.5 more than 7 times higher than long-term median discharge (times per year) Average annual duration of daily discharge being 1.9 6.5 0 1.1 6.7 7 times higher than long-term median discharge (days) Frequency of daily discharge being lower than the 2.2 2.2 2.1 1.5 2.2 long-term median minimum discharge (times per year) Average annual duration of daily discharge being 10.1 6.7 10.1 13.0 6.3 lower than the long-term median minimum discharge (days) Long-term median maximum discharge divided by 6.7 18.8 2.7 5.0 7.7 the median discharge Long-term median minimum discharge divided by 0.38 0.12 0.61 0.70 0.59 the median discharge

Table 2.2 Examples of the different hydrological regime types of Danish streams [1].

Figure 2.6 Trend in annual precipitation at climate stations and annual average runoff in Climate station 16 large Danish streams at gauging stations Gauging station during the recent 75 years. Stars indicate +40 mm signiﬁ cant increase. [3]. +26 mm +35 mm Figure 2.7 Trends in annual average runoff in three streams in Jut- +95 mm* land, Denmark from 1920–1996 shown as 5 years running averages. +69 mm +12 mm The line shows the upward trend in River Gudenå and River Skjern and the downward trend in Århus Stream the latter being inﬂ uenced by water abstraction. +70 mm +40 mm* 600 +125 mm* +7 mm 500 +83 mm* +44 mm +71 mm* 400 +83 mm*

+31 mm +76 mm* (mm) 300 +35 mm +71 mm* +218 mm* -7 mm +33 mm +40 mm +78 mm* 200 +38 mm Runoff +151 mm* 100 +225 mm* 0 1920 1930 1940 1950 1960 1970 1980 1990 2000 Århus Stream River Gudenå River Skjern 2 Hydrology, sediment transport and water chemistry 33

Stream Annual runoff Annual Runoff Difference Change Control period Scenario period in runoff in runoff (1961–1990) (2071–2100) using the HIRHAM model and the (mm/yr) (mm/yr) (mm/yr) (%) IPCCa2 scenario [5]. Table 2.3 shows Odderbæk Brook, 191 226 35 18 N. Jutland the predicted differences between the two 30-year periods in terms of Bolbro Brook, 524 602 78 15 S. Jutland precipitation, temperature, potential Østerbæk Brook, 71 81 10 14 evaporation and runoff parameters for N. Zealand 10 smaller Danish streams situated all Ellebæk Brook, 275 325 50 19 over country. Average annual precipi- W. Jutland tation increases cause a higher aver- Maglemose Stream, 83 90 7 9 age annual runoff because the main Zealand increase in precipitation is in autumn Ølholm Brook, 344 391 47 14 and winter. The predicted climate and E. Jutland hydrological changes will increase Lyby-Grønning Stream, 151 180 29 20 sediment transport as well as delivery N. Jutland of nitrogen and phosphorus to the Lillebæk Brook, 179 214 35 20 streams if the agricultural production Funen and practices remain unchanged [5]. Højvads Rende Stream, 141 188 47 34 Lolland

Sediment transport Bagge Stream, 262 292 30 11 Bornholm The sediment input to Danish streams derives from a variety of sources Table 2.3 Forecast of the impact of climate change on annual and seasonal runoff in Danish among which erosion of stream banks, streams based on downscaled predictions from the climate model HIRHAM. The scenario soil erosion and surface runoff and involves a doubling of CO2 concentrations. sediment delivery via tile drainage pipes have been found to be the most Ripples composed of sand an silt-size sediments are typically found in must Danish lowland important (Box 2.2). The importance streams, as a sign of excess sediment bedload. of stream bank erosion as a sediment source for Danish streams was also documented in a survey of soil and bank erosion along 33 agricultural ﬁ elds in 15 small Danish streams [6]. Transport of suspended sediment in Danish lowland streams is generally low when compared to other regions of the world with mountains and steeper slopes (Box 2.2). However, the transport of suspended sediment varies considerably from stream to stream depending on the hydrological regime. The inter-annual variation in the transport of suspended sediment in a stream is also very high as shown for 14 small streams for the period 1993–2000 where the transport of suspended sediment was measured utilising ﬂ ow-proportional or time-proportional sampling 34 RUNNING WATERS

Box 2.2 Sediment transport in Danish streams

A sediment budget showing the sources of suspended sediment to a small 200 stream in eastern Jutland, Denmark, 150 reveals that although bank erosion is the main source, soil erosion and 100 tile drains also deliver a substantial (kg/ha) proportion of sediment to the stream. 50 The sediment budget will show great

year-to-year variations due to the tight Suspended sediment loss 0 relationship between climate and ero- 0 100 200 300 400 sion. For example, rill erosion on arable Runoff (mm) fi elds in Denmark has been shown to vary 10–20-fold between wet and dry May 1993 to April 1994 May 1994 to April 1995 years. Subsurface tile drainage water 15% 11% Soil erosion and surface runoff 19% 0% Suspended sediment yields in Danish Stream bank and bed erosion 66% 89% streams are low when compared to other regions of the world. The Estimated contribution of subsurface drainage water, surface runoff and stream bank suspended sediment transported in and bed erosion to total suspended sediment export from the 11 km2 Gelbæk Stream Danish streams is very fi ne-grained and catchment in central Jutland, Denmark (adopted from [8]). is therefore of importance as carrier of environmentally harmful adsorbed Stream Catchment Measurement Suspended Sediment substances, such as phosphorus, heavy area period sediment load bed load metals, certain pesticides, PAHs, etc. (km2) (tonnes/km2 (tonnes/km2 The transport of suspended sediments per year) per year) in Danish streams varies considerably Vorgod Stream 455 1976,1977 7.1 < 5.0 from year to year, and the majority of River Skjern 2,220 1976,1977 6.4 12.4 the annual suspended sediment yield River Skjern 2,220 1994,1995 7.9 10.2 is often transported during a few days Grindsted Stream 238 1970,1974,1975 18.7 22.8 with rain or snow melt. Ansager Stream 138 1970,1974,1975 9.5 19.7 Varde Stream 598 1974,1975 9.4 < 1 The movement of coaser sediment on Brede Stream 292 1968 6.2 14.7 the stream bed (bed load) is seldom measured in Danish streams because it Grønå Stream 603 1968,1981 6.8 10.3 is extremely diffi cult to obtain precise River Gudenå 1,787 1974 3.4 < 1 measurements [7]. However, the few Mattrup Stream 88 1974 2.2 < 1 existing quantitative measurements Århus Stream 120 1986–1987 14.3 < 1 show that the bed load is considerably Lyngbygaard Stream 126 1986–1987 28.4 < 1 higher in the dominantly groundwa- Tude Stream 146 1970,1971 5.4 < 1 ter fed streams in western Jutland as Suså Stream (upstream) 42 1978,1979 7.7 < 1 opposed to streams in eastern Jutland Suså Stream 601 1978,1979 5.6 < 1 and on the islands. Average annual suspended sediment load and sediment bed load in different Danish streams [7, 9]. 2 Hydrology, sediment transport and water chemistry 35

programmes (Box 2.2). Such accurate High alkalinity and pH in streams Acidity knowledge on the transport of sus- are associated with calcareous soils in pH pended sediment in streams is valuable the catchments in eastern Jutland and <7.0 as fi ne sediment is an important carrier on the islands where glacial till deposits 7.0–7.5 7.5–8.0 of phosphorus, heavy metals and from the last glaciation period prevail 8.0> micro-pollutants. (Weichsel 14,000 years ago). Such soil No data The bed load of sediment in Danish types have a natural buffering capac- streams has been less intensely stud- ity, reducing the risk of acidifi cation. ied and the majority of information Western Jutland in contrast, has mineral available is related to large streams weathered 140,000 years old moraines in western Denmark with a stable from the Saale glaciation, and coarse hydrological regime and thus a high textural soils on the outwash plains from sediment transport capacity during the Weichsel period. Both soil types are the entire year. Large differences in the low in carbonates and vulnerable to bed load of sediment in Danish streams acidifi cation. Moreover, smaller areas have been measured for the different with windblown deposits along the west Figure 2.8 pH in Danish streams illu- stream types (Box 2.2). coast of Denmark and in some central strated as annual means for streams in 67 Even though sediment transport parts of Jutland are also low in carbon- larger river catchments during the period is low in Denmark compared to most ates and vulnerable to acidifi cation. 1989–1998. other countries, high delivery of sedi- Most Danish streams generally ment to streams from unstable stream have a high or very high buffering – banks causes ecological problems capacity (>0.1 mmol HCO3 /l) against because of accumulation of sand and acidifi cation. A buffering capacity of Ochre – 2+ mud on the stream bed during low-fl ow 0.02–0.1 mmol HCO3 /l can be found Ochre (Fe(OH)3) and ferrous iron (Fe ) periods. Straightening and channelisa- in smaller Danish streams in vulner- can damage the fl ora and fauna in tion of most Danish streams have trig- able areas, whereas a low buffering streams. High concentrations of fer- – gered both a higher sediment input to capacity of 0.01–0.02 mmol HCO3 /l rous iron (>0.5 mg/l) are toxic for fi sh the streams and a lowering of the trans- is rarely found [10]. Thus, the major- and macroinvertebrates, and precipi- port capacity, thereby forcing stream ity of Danish streams have, on aver- tated ochre on spawning gravel, plant managers to excavate streambed sedi- age, an annual pH exceeding 6.5 surfaces and animal gills can physi- ment as part of stream maintenance. and should not suffer any biological cally inhibit their organisms. Ochre The natural physical conditions in damage owing to acidity (Figure 2.8). is ferric (Fe3+) oxy-hydroxides that Danish streams are therefore severely However, streams and lakes of low or precipitate under aerobic conditions impacted. very low buffer capacity have been at pH higher than 4 in surface water documented to show a downward because of the input of soluble ferrous Acidifi cation trend in pH during the period 1977– iron and ferric iron ( pH<3) from soil Acidifi cation because of acid rain or 1989. The strong recent decline in the water and groundwater. Ferrous iron is enhanced nitrifi cation caused by agri- emission of acidifying gases (mainly found naturally in some deep anaero- cultural practice affects sensitive water sulphure oxides) to the atmosphere has bic groundwater aquifers, but the bodies located in carbonate-poor soils decreased the potential problem with main source of ferrous iron in Danish with limited buffer capacity to neutral- acidifi cation of Danish streams. How- streams is from oxidation of natural ise acidic compounds. In areas of low ever, the large emission and deposition deposits of pyrite (FeS2) in waterlogged buffer capacity, acid rain and drainage of ammonia from Danish agriculture is soils. The water level in these soils has water also release aluminium from currently as important as acid precipi- in many cases been lowered due to soils into lakes and streams, generating tation and may locally acidify Danish reclamation of agricultural land. Land toxic conditions for aquatic species. streams of low buffering capacity. drainage has reduced the groundwater 36 RUNNING WATERS

Total iron Fe (mg/l) <0.5 0.5–1.0 1.0–1.5 1.5> No data

Figure 2.9 Concentrations of total iron in Application of slurry from an increasing pig production is now adays done in the spring Danish streams as annual means for streams due to the increase in storage facilities (9 months) demanded by the 1st Action Plan on the in 67 larger river basins during the period Aquatic Environment. 1989–1998.

table and caused oxidation of pyrite Nutrients arable land reached a maximum in the to ferrous iron, sulphate and protons The quantity of nitrogen and phospho- early 1980s, following a strong increase that escape with the drainage water to rus exported to the streams and lakes in the consumption of chemical fertilis- surface waters (see Box 2.1). Deposits in Denmark and their sources has been ers during the 1960s and 1970s (Box of pyrite in Danish soils are mainly calculated for the period 1989–2002 2.4). The consumption of nitrogen found along coastal areas and fl ood- (Box 2.3). The amount of nitrogen and and phosphorus in chemical fertiliser plains in Jutland where old brackish phosphorus discharged to water from has decreased in Denmark since the or marine deposits exist. A survey of point sources has been reduced by 70% mid-1980s due to the implementation low-lying areas in south-western and for nitrogen and 84% for phosphorus of four major Danish Action Plans on northern Jutland showed that 25% of during the same period. Thus, nitrogen the Aquatic Environment. The Action almost 500,000 ha low-lying soils have and phosphorus losses from diffuse Plans have posed several restrictions a potential for releasing ferrous iron sources and the losses from agriculture on Danish farmers, especially regard- and cause ochre problems in surface in particular, have become increasingly ing their storage facilities for liquid waters. The concentration of total iron important. manure, application times of liquid in Danish streams refl ects these condi- Important indicators of the pollu- manure on the soils, demands for uti- tions as the concentrations are highest tion potential from agricultural produc- lisation of nitrogen in animal manure, in western Jutland and lowest on the tion are the consumption of chemical reductions of livestock densities and islands (Figure 2.9). fertiliser and the production of animal reductions of quota for allowed use of manure on arable land. Several sci- nitrogen for crops [12]. entists have documented a positive Nitrogen and phosphorus cycling in relationship between the consumption Danish agriculture has been monitored of chemical fertiliser in agriculture and since 1989 in fi ve small agricultural the resulting concentration or transport catchments by use of questionnaire of nutrients in streams [11]. The nitro- surveys at the fi eld level and measure- gen and phosphorus input to Danish ments of nitrogen and phosphorus in 2 Hydrology, sediment transport and water chemistry 37

Box 2.3 Sources of Nitrogen and Phosphorus

Sources and transport of total nitro- The weather, particularly precipitation, soil water, drainage water, groundwa- gen and total phosphorus through is of great importance for the diffuse ter and stream water [2]. The ﬁ ve catch- Danish streams to coastal waters loadings of nitrogen and phosphorus ments represent typical Danish farming have changed considerably during to Danish coastal waters as can be systems ranging from crop production the period 1989–2003. Discharges of seen from the relationships between on the predominantly loamy soils on nitrogen and phosphorus from point annual nutrient loadings and annual the islands to animal production on sources (especially sewage treatment runoff during the period 1989–2003. the sandy soils in Jutland. The data col- plants) have been greatly reduced and lected can be used to establish overall consequently the diffuse sources have nitrogen and phosphorus budgets for gained in signiﬁ cance during this 14- small catchments, revealing the major year period. input and outputs of nutrients (Box 2.4). Dissolved inorganic nitrogen is responsible for the main transport of nitrogen in Danish streams that drain catchments dominated by agricultural 20,000 – land use (80–90%). Nitrate-N (NO3 -N)

) 16,000 is the main constituent (78–85%). The 3 average annual loss of total nitrogen 12,000 from agricultural land in small catchments was nearby two times higher (million m 8,000 from loamy catchments (23 kg N/ha) 4,000

Runoff than sandy catchments (12 kg N/ha) 0 during the monitoring period 1998– 2003 (Box 2.4). In comparison, the aver- 140,000 age annual loss from predominantly 120,000 forested catchments was 2.3 kg N/ha during the same period. 90,000 (tonnes) Dissolved reactive phosphorus 60,000 (inorganic) constitutes on average 38% of the total phosphorus in streams Nitrogen 30,000 draining loamy and sandy catchments 0 in Denmark. Particulate phosphorus associated with inorganic and organic 10,000 matter is of great importance in Danish 8,000 streams. The speciation of the phos-

(tonnes) phorus export from small agricultural 6,000 catchments implies that the pathways 4,000 are linked partly to the movement of water in the soil column and partly Phosphorus 2,000 to erosional processes. The average 0 annual export of total phosphorus 1981–88 89 90 91 92 93 94 95 96 97 98 99 00 01 02 2003 from loamy (0.53 kg P/ha) and sandy

Direct discharges agricultural catchments (0.51 kg P/ha) Point sources freshwater was almost identical during the moni- Diffuse runoff toring period 1989–1998 (Box 2.4). In comparison, the average annual export 38 RUNNING WATERS

Box 2.4 Nitrogen and Phosphorus losses from Danish agriculture

Nitrogen (1998/99–2002/03) (kg N/ha) Phosphorus (1989/90–1998/99) (kg P/ha)

Manure Fertiliser Manure Fertiliser Manure Fertiliser Manure Fertiliser 52% 28% 36% 52% 64% 36% 39% 61% Other 15% Other 23%

Application Application Application Application 243 183 31.1 26.6 Harvest Harvest Harvest Harvest 147 109 19.1 21.3

12 23 0.51 0.53

Sand Loam Sand Loam 85 53 0.066 0.034

The consumption of nitrogen and phos- rus. Since then a dramatic reduction Large differences are observed in the phorus in chemical fertilisers in Danish has taken place due to the adoption of nitrogen and phosphorus cycling within agriculture increased after World War several Danish Action Plans to reduce two main types of Danish agricultural II and peaked in the early 1980s for nutrient loadings of the aquatic envi- catchments. The nitrogen and phos- nitrogen and mid-1970s for phospho- ronment. phorus surplus is exceedingly higher in sandy catchments in Jutland due to high livestock density and applica- 150 tion of manure. The resulting nitrogen 125 leaching is also highest in the sandy catchments, whereas the resulting river- 100 ine nitrogen transport is highest in the 75 loamy catchments. That higher amounts (kg N/ha) of N escape to streams from loamy 50 catchments than from sandy catch-

Consumption of nitrogen 25 ments – although nitrogen leaching is 0 highest in sandy catchments – must be 1961 72 8586 87 88 89 9091 92 93 94 95 96 97 98 99 00 01 02 2003 ascribed to differences in the hydro- 25 logical cycle of the two catchments types. Thus, a much larger part of net 20 precipitation falling on loamy catchments is rapidly discharged to streams 15 due to inﬁ ltration through macropores

(kg P/ha) 10 to tile drains linked to the streams, as opposed to sandy catchments, 5 where most of the net precipitation

Consumption of phosphorus inﬁ ltrates to groundwater resulting 0 in long residence times. The nitrate 1961 72 8586 87 88 89 9091 92 93 94 95 96 97 98 99 00 01 02 2003 passing the redox zone in groundwater Chemical fertiliser Animal manure is also subdued to denitriﬁ cation due to biogeochemical processes which transform nitrate to gaseous nitrogen 2 Hydrology, sediment transport and water chemistry 39

Organic matter Dissolved BOD (mg/l) orthophosphate PO -P (μg P/l) <1.5 4 (N or N O). In the case of phosphorus, 1.5–2.0 <80 2 2 2.0–2.5 80–120 the sorption potential for dissolved 2.5> 120–160 inorganic phosphorus in the topsoil No data 160> No data and subsoil compartments is important for the amount of phosphorus leached through the soil. Usually only a small part of the phosphorus surplus is lost from the topsoil due to leaching. The soil can, however, be overloaded with phosphorus and for specifi c threshold of phosphorus saturation, depending on soil type, leaching of large quantities of phosphorus will commence. Total phosphorus Total nitrogen This situation has been confi rmed (μg P/l) (mg N/l) in some Danish fi elds, and especially <150 <4 drained organic soils are very vulnera- 150–200 4–6 200–250 6–8 ble to phosphorus leaching because of 250> 8> their much lower phosphorus sorption No data No data capacity. Transport of phosphorus in streams draining agricultural catchments is much higher than normal phosphorus leaching from agricultural soils. Therefore, other hydrological pathways are important for the delivery of phosphorus to streams, mainly due to soil erosion and surface runoff, stream bank erosion and the loss of particulate phosphorus from the soil Figure 2.10 Maps showing the average annual concentration of BOD, total nitrogen, total surface via macropore-fl ow [13]. phosphorus and dissolved reactive phosphorus during the period 1989–1998. The maps are constructed based on measurements in streams within 67 larger basins. The loss of nitrogen and phosphorus from agricultural areas is much higher Compound Number of streams Range of maximum Limit values than background losses as measured in monitored for concentrations for discharges to small streams draining predominantly heavy metals observed in 4 larger Danish freshwater Danish streams forested catchments. Arsen (Ar) 2 1.4–3.1 4 30 0.6 Cadmium (Cd) 3 0.04–0.16 5 Chromium (Cr) 2 2.9–12 10

20 (kg P/ha) 0.4 (kg N/ha) Cobber (Cu) 4 3.3-21 12 Mercury (Hg) 1 0.008* 1 10 0.2 Lead (Pb) 1 7.0* 3.2

Nitrogen loss 0 0 Nickel (Ni) 5 2.9–34 160 Phosphorus loss Zink (Zn) 4 18–81 110 Natural or forested areas Agricultural land Table 2.4 Concentration of heavy metals in larger Danish streams draining mixed agricultural and urban catchments (data from the year of 2000). 40 RUNNING WATERS

Box 2.5 Pesticides

The consumption and application of pesticides in Danish agriculture has been was 0.079 kg P/ha from undisturbed reduced since the late 1980s. catchments. A statistical analysis of changes in ﬂ ow-adjusted nitrogen concentra- 4,000 Herbicides tions in 63 Danish streams draining Fungicides Insecticides agricultural catchments without major 3,000 point sources, has shown a general downward annual trend of 27% during 2,000 the period 1989–2002. The downward trend was higher in loamy catchments 1,000 than in sandy catchment because of

Pesticide consumption higher livestock densities and longer (tonnes active ingredients) 0 residence times in groundwater in 1992 93 94 95 96 97 98 99 00 01 02 2003 sandy areas [12]. A similar statistical trend analysis has not shown signifi - Many pesticides are detected in Danish streams and normally there is a rela- cant changes in phosphorus concentra- tionship between amounts of active substances applied and the frequency of tions in 38 smaller Danish streams in detection in the streams. Herbicides are generally detected more often than agricultural catchments. fungicides and insecticides, as many of these pesticides are more hydrophobic In spite of successful adoption of and are thus sorbed in the soils or in stream sediments. Fungicides and insecti- several Action Plans for combating cides are detected in stream sediments and sometimes in very high concentra- nutrient pollution from point sources tions that are damaging to the fauna [14]. The concentration of pesticides in and diffuse sources, the concentrations Danish streams is very low and usually below the detection limits (1–10 ng/l). and loads of nitrogen and phosphorus However, high concentrations of newly applied pesticides are occasionally in Danish streams are still very high, detected in small streams draining agricultural catchments, especially during causing eutrophication of many water spates where pesticides are transported by surface runoff or macropore-fl ow bodies (Figure 2.10). directly to streams. Heavy metals and organic Detection of pesticides in 25 smaller Danish streams draining dominantly agricultural micropollutants catchments during the year of 2000. (Number of water samples analysed = 189) One of the fi rst major pollution stories

Compound Detection Average Median Maximum hitting the Danish newspaper headlines frequency concentration concentration concentration in the late 1960s was the ﬁ nding of (%) (ng/l) (ng/l) (ng/l) excessive amounts of mercury in ﬁ sh Glyphosate (H) 73.9 132 77 1,400 and sediment in the Varde Stream in Dimethoate (I) 2.1 46 27 120 south-western Jutland. The mercury Bentazone (H) 37.0 61 20 1,200 Ioxynil (H) 5.9 75 20 440 pollution was caused by discharge of Pirimicarb (I) 3.8 23 26 42 sewage from an industrial plant using Fenpropimorph (F) 1.6 49 27 110 inorganic mercury in their produc- Ethofumesate (H) 5.4 187 65 920 Diuron (H) 23.8 57 25 360 tion of chemicals for use in industrial Terbuthylazine (H) 32.3 72 29 1,260 products. Heavy metals are now Propiconazole (F) 6.4 138 20 1,400 generally detected in low concentra- Pendimethalin (H) 11.1 75 38 580 Esfenvalerate (I) 0 - - - tions in Danish rivers and do not create immediate concern about their ecologi- (H) = Herbicide; (F) = Fungicide; (I) = Insecticide. cal impacts (Table 2.4). The reason is *: Only 62 water samples analysed for in the case of esfenvalerate. the strong decrease in the use of most poisoning heavy metals, such as mer- 2 Hydrology, sediment transport and water chemistry 41

cury, cadmium and lead in households, regulation of pesticide application on pollution by increasing the nutrient industrial production and chemical agricultural land has certainly helped retention capacity in the landscape. fertilisers. Moreover, the introduction to reduce pesticide loss to surface The main measures are rehabilitation of phosphorus precipitation in most waters in Denmark. of former riparian wetlands and wet Danish sewage treatment plants has The presence of other organic meadows by closing down drain- increased the treatment effi ciency for micro-pollutants has been measured in age schemes where surface water heavy metals by more than 50%. The fi ve larger Danish streams since 1998 is pumped away, re-meandering of concentrations of heavy metals in outlet (Table 2.5). Among the most frequently streams to re-establish the natural water from sewage treatment plants are detected substances are PAHs (Poly- hydrological interaction between the currently below the stipulated qual- cyclic aromatic hydrocarbons) derived stream channels and their fl oodplain, ity objectives required for the aquatic from exhaust gas from cars and other recreation of former lakes, and estab- environment. However, there are still combustion processes. Among the lishment of uncultivated buffer zones some doubts about the importance most discussed harmful substances are along watercourses. The philoso- and fate of nickel, zinc and copper in those with hormone disturbing effects, phy is to regain some of the natural animal manure, which is applied in for example nonylphenols, DEHP, nutrient sinks in the landscape through great quantities to agricultural areas biphenol A and natural estrogens in the denitrifi cation of nitrate, sedimenta- and thus can be transported to streams. aquatic environment. tion of organic nitrogen and particulate Moreover, the presence of naturally Feminisation ( intersex) of fi sh was phosphorus and sorption of dissolved high concentrations of heavy metals in found for the fi rst time in watercourses phosphorus in the rehabilitated areas. deeper, reduced groundwater (arsenic) near the city of Århus, Jutland [15]. Research has proven that the effi ciency or in groundwater where water abstrac- This phenonenon is probably caused of natural systems to remove nitrogen tion has lowered the groundwater table by natural hormones discharged with and retain nitrogen and phosphorus is (nickel) could locally pollute surface domestic sewage into the aquatic very high in some natural systems (Box water with heavy metals where deeper environment and to a lesser extent by 2.6). groundwater enters the streams. hormones originating from the syn- Improved nutrient managing at the A major reduction in the use of thetic substances. source (farmer) can be supported by pesticides in Danish agriculture has measures to rehabilitate the natural been documented (Box 2.5). The Regaining the natural retention conditions in catchments by, for exam- decrease is mainly due to the introduc- processes in river basins ple, reinstating naturally vegetated tion of new modern pesticides that are The last two Danish Action Plans buffer zones along river channels, applied in much smaller quantities (1998 and 2003) for reducing nutrient restoring river channels and fl ood- than old pesticides. The occurrence and pollution of the aquatic environment plains and reinstating wetlands and concentration of pesticides in streams involved measures to combat nutrient lakes (Box 2.7). have been monitored every year since 1999 in 25 streams draining agricultural catchments. The occurrence and concentration of the most important pesticides detected are shown in Box Environmental Number of Mean values Max. values pollutants detections (μg/l) (μg/l) 2.5. Even though some pesticides and Trichlor-ethylene 12 0.285 1 metabolites are detected frequently, the concentrations of pesticides in streams Acenaphthene 6 0.0085 0.025 are low. During certain periods of the Table 2.5 Concen- Flouranthene 7 0.0105 0.033 year (spraying season with either low trations of organic Pyrene 8 0.0115 0.031 fl ow or spates) the maximum concen- micropollutants Linear alkyl benzene 6 1.75 14 trations of certain pesticides surpass detected most often sulphonate (LAS) the threshold concentration. Strict in Danish streams. 42 RUNNING WATERS

Box 2.6 Nutrient retention

Undrained riparian areas can remove m.a.s.l. substantial amounts of nitrate through 1,080 kg NO3 denitriﬁ cation whereby – under 48 anaerobic conditions through the use 47 Stream Stream of organic carbon (biologically) or Peat Peat Tiledrain pyrite (chemically) – nitrate is trans- 46 1,100 kg NO3 20 kg NO3 Loamy sand formed to gaseous N. If the riparian Loamy sand 45 area is tile-drained instead nitrate will Gyttje Gyttje pass through the area without being 44 0 1020304050607080 m 0 1020304050607080 m reduced.

Wetlands Streams Lakes Estuaries The removal of nitrate via denitriﬁ ca- 32–2,100 50–700 65–220 20–550 tion is a process that takes place under kg N/ha per year kg N/ha per year kg N/ha per year kg N/ha per year anaerobic conditions in all soils and water compartments where nitrate exists.

Soil erosion Soil erosion 1,210 kg 7,230 kg Sedimentation Sedimentation 430 kg 780 kg 4,470 kg Wide non-cultivated buffer zones 0 kg along stream channels can function as 0 kg 2,760 kg an effective ﬁ lter for nutrient-rich soil Buffer zone Field Sloping particles mobilised through soil erosion 29 m Downhill field processes and transported with surface Buffer zone 0.5 m runoff towards the streams.

Soil erosion can deliver huge quantites of sediment and phosphorus to aquatic ecosystems. 2 Hydrology, sediment transport and water chemistry 43

Box 2.7 Restoring rivers and ﬂ oodplains

Perspectives and future threats Rehabilitation of river and fl oodplains is one of the important management We have gained great insight into the schemes that can be used by river basin managers to reduce diffuse nutrient hydrological and chemical conditions pollution. The effect of rehabilitating streams, wetlands and lakes in a 100 km2 in Danish streams during the last 15–20 catchment in Jutland, Denmark was simulated by the TRANS model [16]. The years. Research has helped us to under- effect measured as increases in nitrogen removal and phosphorus storage in stand many of the processes governing streams, lakes and fl oodplains following rehabilitation, was very high. the hydrochemistry of streams and catchments, especially the threats from acidifi cation, ochre and nitrogen pollution. However, we now face a new era where treatment of point source discharges alone cannot solve the problem of achieving good ecological quality surface waters as demanded in the EU Water Framework Directive. Thus, we need to improve the understanding of physical processes that regulate transport of substances from agricultural fi elds to streams, and biogeochemical processes that remove, transform or store substances during the transfer from land to sea. This information is required by managers in order to target pollution problems in surface waters and it is necessary for preparing new Nitrogen removal (tonnes) Phosphorus storage (tonnes) and better hydrochemical models to Pre-restoration Post-restauration Pre- restoration Post-restauration predict possible future changes in the Watercourses 5.2 21.6 0.24 –0.08 aquatic environment of regional man- Lakes 33.5 28.0 –1.39 0.09 agement schemes and climate change. Floodplains 0.13 31.3 0.35 0.89 44 RUNNING WATERS

The springbrook is the ﬁ rst part of the river system. 45

3 From spring to river – patterns and mechanisms A river system consists of a branched network of streams, but how do environmental conditions and the distribution of organisms vary along the system? The depth, width, discharge and mean age of the water increase from source to mouth and the physical and chemical characteristics of the water gradually become more infl uenced by contact with the atmosphere and in-stream processes. Species composition and diversity also change along the stream. These patterns are infl uenced by the location and site-specifi c characteristics of the reaches. In addition, they are infl uenced by the dispersal of organisms between reaches making the biota at each site dependent on the biota at neighbouring sites.

Streams – a network with Flow and substrata from spring unidirectional ﬂ ow to river As described in chapter 1, streams With increasing distance from the form a unique ecosystem because spring, the stream usually becomes of the linear connection in the net- wider, deeper and transports more work, the unidirectional ﬂ ow and the water (Figure 3.1 and 3.2). Slope, tem-

intimate contact with the land. The perature and CO2-concentration also network generates a ﬁ nely branched change downstream. connection transporting water and Along the course, the slope matter from the terrestrial areas to the decreases [2]. Denmark has no moun- sea. These aspects are incorporated in tains, however, and its streams follow models for transport of organic matter more irregular courses, shifting and nutrients and for transformation between reaches of steep or low slopes, of oxygen and carbon, and they are interspersed with stagnant lakes with integrated in the River Continuum long water retention. Nonetheless, the Concept (Box 3.1). steepest slopes are usually found in Species composition and richness at the springs and brooks, and the lowest the individual stream reaches are often slopes in rivers close to their outlet treated as if they were independent [4, 6]. However, due to irregularities, of conditions at neighbouring reaches reaches with steep slopes and high even though organisms are extensively physical stress may occur anywhere dispersed throughout the stream along the course of a river. network. A comparison of 208 Danish stream reaches during summer showed a doubling of the mean water velocity from 46 RUNNING WATERS

Box 3.1 River Continuum Concept (RCC)

The RCC was formulated by American stream ecologists to The original North American descriptions of gradual changes conceptualise general patterns of input and conversion of in the photosynthesis to respiration ratio from very low organic matter along the stream course [1]. Some of the values in the upper sections (P/R << 1), highest values in the original concepts were able to withstand later tests while mid-sections (approaching 1) and lower values in down- others had to be abandoned [2, 3]. stream sections (e.g. 0.2–0.8) presuppose that the upper sections are highly shaded by forests, the mid-sections are Streams receive large amounts of terrestrial organic matter unshaded and that photosynthesis in the lower sections is along the entire course from their source to their mouth. constrained by light attenuation in deep water. Thus, streams usually decompose more organic matter than they produce by photosynthesis. The net balance of the Recent studies have revealed that organic inputs from the system metabolism in streams is said to be heterotrophic fl oodplain in the lower sections had been underestimated, because the combined annual respiration of all organisms especially during fl ooding [2]. Neither has there been any exceeds the photosynthesis of algae and plants. The ratio verifi cation of two other RCC-predictions: that there is a sys- of annual photosynthesis to annual respiration (P/R) may tematic downstream increase in the relative quantity of fi ne be close to 1 in desert streams because they receive little particulate organic matter relative to coarse organic material terrestrial material. In spring, when irradiance is high and (e.g. leaves), due to the processing of material by inverte- bank shading low and while the terrestrial input of organic brates, and a concomitant decline of large shredders (e.g. matter is low and respiratory degradation is restricted by low trichopterans) [3]. Coarse organic material is often added to temperatures, photosynthesis may also exceed system respira- the stream along the entire course without a consequent, tion in small Danish streams [4]. On an annual basis, however, predictable change in the proportion of shredders. The system respiration surpasses photosynthesis at all reaches in proportion of coarse material is not systematically higher in Danish streams and in most other streams worldwide. forest streams than in open streams [5]. Shredders do not solely consume shredded terrestrial leaves, as they may also The original RCC proposed systematic changes in photosyn- graze on live submerged plants and macro-algae, and occa- thesis, respiration and the ratio between these along the sionally even act as predators. stream course but both processes depend on local organic inputs, water depth and turbidity, stream bed stability and Despite the criticism of the RCC, it remains useful to consider terrestrial shading. streams as a continuum by tracking matter and processes as the water moves downstream. It requires, however, integra- tion of local sources of organic matter and determinants of photosynthesis and decomposition in order to achieve an accurate overall picture.

0.13 m/s in streams smaller than 2 m plant patches and behind bends close regulated state. In a nationwide study to 0.26 m/s in streams wider than 6 m to the banks [2, 4]. of 40,000 quadrate (25×25 cm) in 208 (Figure 3.2). This difference has conse- Local fl ow conditions above the reaches in 45 streams, we did not quences for the plants directly exposed stream bed regulate erosion and observe systematic changes in the main to the free-fl owing water. Flow condi- sedimentation. Although intuitively substrata along the stream courses. tions are different for micro-organisms, we may believe that steep headwaters Sand was the most common bottom invertebrates and fi sh that live in close have coarse gravel substrata while substratum in both small and large contact with the stream bed or the large rivers have fi ne-grained mud, streams. banks. Low fl ow velocities may prevail this does not apply to Danish low- We also determined substratum immediately above the stream bed at land streams – neither in their origi- heterogeneity by calculating how often deep sites and in the shelter behind nal unregulated nor in their current, two neighbouring quadrates had dif- 3 From spring to river – patterns and mechanisms 47

e Light 100 0.75 Light availability in streams is d determined by solar radiation, bank c 10 0.50 shading, surface reﬂ ection and light b attenuation through the water. In small (m)

No. of reaches 0.25 1 a narrow streams, the banks and ter- Mean depth restrial herbs can intercept most of the 10,000 0 incident light. In wide rivers, however, 1,000 large trees are needed to provide the c same shading. 100 0.3 (l/s) bc We can all recall the sight of grass-

Discharge 10 b green water downstream of plankton- 0.2 b 1 rich lakes and the sight of yellowish a

(m/s) water rich in suspended particles 10 0.1 during peak ﬂ ow. Under these cir- 8 Mean velocity cumstances, light attenuation through 6 0 4 the water column is so profound (days) 2 that plants and microscopic algae are 0 1.5 unable to photosynthesise and grow at Mean age of water b b 12345 the stream bottom. These extreme con- Stream order 1.0 ab ditions, however, are not representative a a of the general state of light conditions

Figure 3.1 Three idealised graphs illustra- (1/m) 0.5 in streams, which are much better in ting how large number of small headwater coefficient most instances. streams gradually merge into fewer streams Light attenuation The survey of 208 Danish stream 0 of higher order (top), higher discharge reaches found that during summer, (middle) and higher mean age of the trans- light attenuation through the water of 18 ported water (bottom). d Danish streams was lower at small and 16 large reaches than at reaches of inter- cc mediate size (Figure 3.2). These differ- 14 ences were mainly due to variation in (˚C) b the concentrations of mineral particles,

Temperature Temperature 12 a which were lower in small streams 10 receiving substantial proportions of 0–2 2–4 4–6 6–8 >8 particle-free groundwater and in larger ferent substrata (Box 3.2). Substratum Stream width (m) streams where sedimentation exceeded heterogeneity could be important for re-suspension during low-fl ow periods. the diversity of plants and inverte- Figure 3.2 There are systematic changes in The light attenuation coeffi cient (k) brates because species differ in their mean depth, velocity, light attenuation and describes how rapidly light intensity substratum preferences. However, we temperature along the river continuum from decreases with water depth accord- observed only small and insignifi cant source to outlet. Summer measurements in ing to: Iz = Io exp(-kz), where Io is light differences in mean values of substra- 208 Danish stream reaches illustrate the pat- intensity at the surface, Iz is light inten- tum heterogeneity among different size terns. Columns sharing a common letter are sity at depth z, and exp is the exponen- categories of streams (Figure 3.3). not signifi cantly different from each other. tial function. The mean exponential [Sand-Jensen and Riis, unpubl.]. coeffi cient (k) ranged from 0.7 to 1.3/m among Danish streams of different 48 RUNNING WATERS

Box 3.2 The heterogeneity of the stream bed

The heterogeneity of the stream bed is calculated on the basis of the following 0.8 expression

0.6 where SH is the substratum heterogeneity, i is the transect and j is the quad- 0.4 rat. The value is a number between 0 and 1 and the closer the value is to I,

Heterogeneity 0.2 the higher is the heterogeneity of the 0 stream bed. 0.3 1 3 10 30 Stream width (m)

Figure 3.3 The heterogeneity of the stream bed (Box 3.2) is highly variable among reaches, but does not change systematically along the course of Danish streams [Sand- Jensen and Riis, unpubl.].

width (Figure 3.2). Consequently, an depths allow better heat exchange with a 5.6-fold supersaturation relative to average of between 50% and 27% of the atmosphere than deep water. In the atmospheric equilibrium. Only reaches the light intensity passing the water lower reaches, temperatures are higher located immediately downstream of surface reaches a depth of 1 m [4]. and close to the mean air temperature lakes were close to atmospheric equi- Some groundwater-fed brooks had because the water has been transported librium because of the long retention light attenuation coeffi cients of only for a long time and has come close time and photosynthesis in the lakes. 0.2–0.4/m, which is less than for most to equilibrium with the air. It is well Studies of the long River Gudenå, Danish clearwater lakes. The compari- known that water temperature is very which is interspersed with several sons show that overall, streams have important for the distribution and lakes, demonstrated that there are clearer water during summer than the metabolism of invertebrates and fi sh distinct increases in temperature and mostly plankton-rich Danish lakes, [2, 7], but its infl uence on plants has phytoplankton biomass and decreases which have attenuation coeffi cients of been neglected so far (see later). in CO2 concentration downstream of typically 0.4–4.0/m. Temperature and CO2 concentra- the lakes (Figure 3.4). tion tend to follow a parallel course

Temperature and CO2 in streams (Figure 3.4). Cold water Dispersal of stream organisms

In the same survey, midday tempera- is “young” water rich in CO2 from There is a very good reason why the tures averaged about 11°C for the soil respiration, while warm water is connectivity of the stream network is upper stream reaches and 17°C for the “old” water having lost part of the CO2 important for animals and vegetation. lower reaches during summer (Figure supersaturation due to contact with the Streams are often exposed to consid- 3.2). There are two reasons for this air and photosynthesis. Nonetheless, erable disturbances (e.g. peak ﬂ ows) difference. The small upper reaches are most streams maintain CO2 supersatu- leading to a signiﬁ cant decline or eradi- characterised by “young” water that ration irrespective of stream type and cation of populations at certain reaches, has just recently been received from the location. Thus, 56 stream reaches in so immigration and recolonisation of cold groundwater aquifers. Headwater North Zealand had median CO2 con- new individuals are crucial in main- streams are also more shaded due to centrations of 176 μmol/l in the morn- taining species richness. The protection the narrow channel. Consequently, ing and 90 μmol/l in the afternoon, of species against extinction in streams the water is colder, although shallow corresponding to about an 11-fold and depends largely on the ability of a few 3 From spring to river – patterns and mechanisms 49

populations to survive, even following serious disturbances. 160 Morning From these refugia, species can disperse and re-colonise the entire stream network. 120 (μmol/l) Hildrew and Townsend have described how invertebrates undergo a continuous redistribution in streams [8]. 80 As a result, invertebrate populations are re-established

40 relatively rapidly at reaches that have been subjected to -concentration

2 catastrophic disturbance due to pollution or sediment ero- Afternoon CO 0 sion [3]. It is, therefore, surprising that the assumption of independent behaviour of populations of reaches is main- 160 tained in many biological studies. Chlorophyll Different species do not have equal ability to disperse 120 upstream and downstream. The ability of the species to (μg/l) (%-coverage) become established in new habitats or re-colonise former 80 40 habitats depends on the location of the surviving populations and the biological characteristics of the species. 40 20

Chlorophyll Chlorophyll All organisms are able to disperse passively with the Plants current and consequently the downstream dispersal capac- 0 0

Submerged plants ity usually exceeds the upstream. Plants are able to deliver a continuous supply of seeds, detached shoots and some 25 vegetative dispersal units (e.g. short shoots) to downstream 20 reaches, thereby enhancing the chances of establishing new

(˚C) plant stands or maintaining old stands there. 15 Surprisingly few studies have attempted to quantify the 10 transport of plant dispersal units. Moreover, the dispersal has rarely been evaluated on the basis of the actual distribu-

Temperature 5 tion of plant species. In a study of Ranunculus lingua growing 0 in wet soils, with most of the shoot exposed to air, shoots were tagged and later recovered at a considerable distance Headwater further downstream [9]. A lake, on the stream network, Weir formed a trap for further downstream dispersal. Once lake Lake Mossø Lake Ry Mølle populations have been established, however, not only do Lake Juel they ensure downstream dispersal, but they also enrich the Lake Silkeborg Langsø downstream reaches with many individuals and species, Lake Tange provided that the lake populations are rich and abundant. Estuaurine One hundred years ago, most Danish lakes had clear water and dense, submerged vegetation capable of enrich- Figure 3.4 River Gudenå has many weirs and interspersed lakes. ing downstream reaches with species (Figure 3.5). Unrooted

CO2 concentrations and macrophyte cover decline downstream Ceratophyllum demersum (rigid hornwort), for example, was during summer, while there is an increase in phytoplankton biomass common in outlets from lakes while nowadays this species and temperature. is no longer observed in streams [10]. Also Myriophyllum spicatum and M. verticillatum (spiked and whorled water- milfoil), Ranunuculus circinatus (fan-leaved water-crowfoot) and several large species of Potamogeton (pondweeds) were previously more common. Due to eutrophication, lakes interspersed within streams have entirely lost their sub- 50 RUNNING WATERS

4 Number of species

0 River River River Suså Århus Gudenå

The quality of the lake water inﬂ uences biological communities in the stream below. 100 years ago Present

Figure 3.5 The number of submerged plant merged vegetation. They form a trap Thus, if species richness rises as species has declined markedly at reaches for the dispersal of species between reaches increase in order and width downstream of lakes over the past century. upstream and downstream reaches and downstream, it may not solely refl ect The lakes have been enriched with nutrients they export turbid plankton-rich lake the greater size and more varied that stimulate phytoplankton production water. As a result, there are few, if any environmental conditions [1]. Species and lead to loss of submerged rooted vege- submerged species downstream of the enrichment may also result from the tation. The lakes now export turbid water lakes (Figure 3.4 and 3.5). It has thereby better supply of species and individu- and prevent plant dispersal. The fi gures become more diffi cult for submerged als from upper stream reaches. are from the Suså Stream downstream of species to disperse between lakes and The downstream dispersal is Lake Tystrup-Bavelse, the Århus Stream streams, and this has probably con- sometimes clearly refl ected in the downstream of Lake Brabrand, and the tributed to the development of greater distribution of plant species. For River Gudenå downstream of Lake Silkeborg differences between the submerged example, observations may show that Langsø [10]. vegetation of lakes and streams today two upstream reaches have very dif- than there were 100 years ago [10]. ferent species compositions while the There are numerous small brooks in downstream reach has several species the upper part of the stream network, in common with the two upstream 100 which is richly branched throughout reaches after their confl uence [11]. 80 the landscape. The number of stream sections decline with increasing stream Plant distribution in stream systems 60 (%)

order [2]. Assuming, for example, In a comprehensive study of plant Cover 40 an average 3-fold decrease of stream distribution at 208 unshaded reaches reaches per order, a fi fth order stream of 45 Danish streams, we attempted to 20 system may include 81 small fi rst order answer three important questions: 0 brooks and one fi fth order river. The • How great is the plant cover? 0 5 10 15 20 25 30 combination of the high number of • Which species and plant types Stream width (m) brooks, their varied conditions and the dominate? downstream transport of water is likely • How does species richness vary Figure 3.6 Mean plant cover in the middle- to facilitate the enrichment of species among the reaches? part of the stream bed declines along the along the downstream course of the stream course as plants receive insuffi cient river system. light with increasing depth. 3 From spring to river – patterns and mechanisms 51

50 25

40 20

30 15

(%) 20 (%) 10

10 5 Relative frequency Relative frequency Relative frequency

0 0 0–3 3–6 6–9 >9 0–3 3–6 6–9 >9 Stream width (m) Stream width (m)

Terrestrial submerged plants Water starwort The large species Potamogeton praelongus Amphibious plants Water crowfoot species (long-stalked pondweed) is most common in Obligate submerged plants Waterweed large streams. Pondweed species

Figure 3.7 The mean relative abundance of Figure 3.8 Among the obligate submerged three groups of plants (terrestrial, amphibi- plants there is a change in the relative abun- abundant in small upper reaches not ous and obligate submerged plants) in the dance of species and species groups with wider than 3 m. Ranunculus species aquatic vegetation of 208 Danish stream increasing stream size. Callitriche species such as R. fl uitans (river water-crow- reaches covering a range of widths. Abun- (water-starwort): mainly C. cophocarpa and foot) and R. peltatus (pond water-crow- dance is based on the presence of plants in C. platycarpa. Ranunculus species (water foot) were most abundant in streams about 200 quadrats per reach. Note the gra- crowfoot): mainly R. baudotii, R. aquatile smaller than 9 m. Elodea canadensis dual substitution of terrestrial and amphi- and R. peltatus and Potamogeton species (Canadian waterweed) was most bious plants mainly associated with stream (pondweeds): P. pectinatus, P. crispus, P. per- abundant in 3–9 m wide mid-section banks by obligate submerged plants, as the foliatus, P. natans, P. praelongus, P. lucens, P. streams. Finally, the collection of large width and depth of the reaches increase. berchtoldii and P. frisii in order of decreasing Potamogeton species, which includes Terrestrial plants include species that may abundance. P. crispus, P. natans, P. pectinatus, P. survive, but with poor growth, under water. perfoliatus and P. praelongus (curled, Amphibious plants grow on land as well broad-leaved, fennel, perfoliate and as under water, while obligate submerged long-stalked pondweed), was most plants are usually unable to survive in air. abundant in wide, lower reaches. The distribution of plants is probably infl uenced by changes in tem- The fi rst two questions are discussed tuted a decreasing proportion of the perature and water depth. Growth below, while the third question is vegetation, as reaches become wider experiments in the laboratory have treated in the following section. The and deeper (Figure 3.7). Truly sub- shown that the aforementioned species results showed that the vegetation merged plants, which always grow of Callitriche and Ranunculus are less often reached a very high cover in under water, were the most widely sensitive to temperature variations than shallow Danish lowland streams distributed and abundant at all reaches Elodea and Potamogeton. Callitriche often during summer. Moreover, plant cover forming 45–50% of the cover. has the greatest abundance relative to declined with increasing stream width The relative abundance of differ- the other submerged species during the (Figure 3.6) simply due to insuffi cient ent permanently submerged species winter. The peak abundance of Ranun- light in the deeper areas of the lower changed along the stream course culus is attained during the spring reaches. Not surprisingly, terrestrial (Figure 3.8). Species such as Callitriche and early summer, while Elodea and plants, with sporadic aquatic occur- platycarpa (various-leaved water-star- Potamogeton peak during mid-summer, rence, and amphibious plants consti- wort) and C. cophocarpa were most refl ecting higher temperature require- 52 RUNNING WATERS

that species richness of all insects increases from upper to lower reaches with similar terrestrial surroundings and pollution level [5]. Other studies have shown that species richness of fi sh increases with the size of stream systems (Figure 3.9). An important goal is to achieve a better understanding of the regulation of species richness among plants, invertebrates and fi sh. Differences and similarities among groups of organisms should reveal the importance of dispersal capability and environmental Unshaded lowland streams of medium A shaded forest stream without macrophyte requirements for the distribution of the size usually have very varied submerged growth. different species. Evaluations should vegetation. be performed for individual reaches as well as entire stream systems. In ment [10]. Species of Callitriche are There is a logical explanation for individual reaches, evaluations must small and delicate and have a special this pattern. In very narrow and shal- take into account their surface area, ability to sprout and form emergent low streams there is limited space and habitat heterogeneity, environmental shoots on wet soils. Accordingly, they variability in water depth compared conditions and location. Downstream are considered close to the transition with the conditions in larger streams. reaches of the same length can support from amphibious to submerged plants. Immigration of species from other more species because water tem- Like the amphibious species, most reaches is also less likely to occur in the perature, surface area, depth variation Callitriche species are totally dependent upper than in the lower reaches. and the immigration of species are on CO2, which is present at its highest higher than in upstream reaches. This concentration in cold headwaters. At Species richness of animals in environmental complexity means that the opposite extreme of the spectrum, streams it is diffi cult or impossible to identify in deep, warm, lower reaches, the large While mean species richness of plants the mechanisms that are particularly Potamogeton species benefi t from the remains relatively constant along important for the observed differences higher temperatures, their ability to use the course of Danish streams, the in species richness. bicarbonate, their greater plant size and animals exhibit considerable differ- Relationships between species their ability to form a canopy close to ences. Wiberg-Larsen conducted an richness and surface area for lakes and the water surface [12]. interesting study of species richness terrestrial habitats usually follow an of caddisfl ies from a large number of empirical power equation: S = cAz (i.e. Species richness of plants along Danish stream reaches [6]. Caddisfl ies log S = log c + z log A). The increase streams increased markedly in species rich- rate (z) of local species richness with Total species richness in open Danish ness with increasing width of stream surface area is usually between 0.1 and streams changes very little with loca- reaches. The increase was approxi- 0.5 with a mean value of about 0.3. This tion and width of the stream once it is mately linear when species richness value indicates that species richness in wider than 3 m, where mean species was plotted as a function of stream a local habitat doubles with a 10-fold richness is 12–13 per reach. However, width on logarithmic axes, although increase in the surface area of the habi- mean species richness is consistently the rate of increase in species richness tat. However, it is diffi cult to interpret lower (10 species) in very narrow tended to decline in the widest reaches. the cause of different z-values, and reaches not wider than 3 m. Jacobsen and Friberg have also shown it is particularly diffi cult to evaluate 3 From spring to river – patterns and mechanisms 53

z-values for stream reaches within area, it remains diffi cult to determine to species pool for all stream systems, it stream systems as species richness at what extent an increase in the number is possible to evaluate the infl uence individual reaches can hardly be con- of individuals and increased habitat of stream size, habitat variability and sidered as independent observations. variability are responsible for the water chemistry on species richness of Furthermore, the downstream increase increased species richness [13]. Time is the individual streams. In other cases in water temperature makes it prob- another crucial determinant of species it is possible to evaluate the infl uence lematic to compare the relationship richness. Over thousands and millions of time and immigration history by between species richness and width in of years, older stream systems will tend comparing streams of similar size, streams with the relationship between to have accumulated (via immigra- habitat variability and water chemistry. species richness (S) and surface area tion) a greater number of species than It is important to remember, however, (A) as applied for lakes and terrestrial younger systems. If large stream sys- that history may have a strong impact habitats [13]. tems are also older than small stream on regional differences within one Some of these problems of inter- systems, which is often the case, this continent, and even within one country pretation do not occur when species connection between size and age will [2, 3]. In Denmark, for example, the richness of entire stream systems is tend to strengthen the positive rela- species richness of caddisfl ies is lower evaluated as a function of their size. tionship between species richness and on the main islands of Zealand, Funen Size can be as expressed as either the stream size. and Bornholm than in Jutland (the entire surface area of the stream bottom, In order to learn more about the peninsula connected to the European the water discharge at the outlet or the regulation of species richness it is continent) [6]. As the local species catchment surface area in regions with important to analyse why certain richness derives from the regional similar hydrology (Figure 3.9). Here, stream systems have more species, species pool, there are fewer species in total species richness usually increases while others have fewer species, than local streams when the regional spe- linearly with the size of the stream would be expected from an average cies pool is low. There are also distinct system when plotted on logarithmic relationship, based on all of the stream differences in species composition of axes [14, 15]. An increase in the surface systems examined. In other words, we invertebrates and fi sh over distances of area of stream systems has the impor- need to test the reasons for positive and only a few km between the old streams tant effect of increasing the number of negative variations (residuals) from the (c. 150,000 years) in western Jutland, individuals and thereby the number of main average pattern. If the histori- derived from the glaciers during the different species. However, as habitat cal infl uence of immigration can be last glaciation, and the young streams variability usually rises with system discarded in a region with a common (c. 10,000 years) in eastern Jutland, which were created when the glaciers retreated. Figure 3.9 Examples of increasing numbers of species with increasing scale of the stream Over time there is no distinct upper systems. (A) Species richness of stream invertebrates on Bornholm Island. (B) Species rich- limit for the size of the local and ness of fi sh in larger stream systems. regional species pool. Given time, area and suitable environmental condi- 100 1,000 AB tions then species richness will rise. Unfortunately, the profound reduction 100 of the population size of many stream 10 organisms during the last 100 years, 10 together with the ever-rising human populations and the more intense

Number of species 1 1 use of land and water throughout the 50 500 5,000 10 100 1,000 10,000 100,000 world, is likely to even further reduce Area of streambed (m2) Catchment area (km2) species richness of the streams over the coming century. 54 RUNNING WATERS

Streams are characterised by large spatial variability in ﬂ ow. 55

4 Water fl ow at all scales Continuous water fl ow is a unique feature of streams and distinguishes them from all other ecosystems. The main fl ow is always downstream but it varies in time and space and can be diffi cult to measure and describe. The interest of hydrolo- gists, geologists, biologists and farmers in water fl ow, and its physical impact, depends on whether the main focus is on the entire stream system, the adjacent fi elds, the individual reaches or the habitats of different species. It is important to learn how to manage fl ow at all scales, in order to understand the ecology of streams and the biology of their inhabitants.

Macroscale and overview ance to water fl ow is reduced during Flow conditions in stream reaches are contact with the streambed and the often described by measurements of banks. Finally, submerged plants resist discharge, water velocity and water water movement so that mean velocity stage. If the downstream slope of declines with increasing plant biomass. the water surface, the cross section Moving water encounters a greater profi le and the composition of the resistance in a meandering stream, stream bed are also known, empiri- which reduces fl ow velocity compared cal equations can be used to estimate with a straight channel. This is why hydraulic resistance against the fl ow channelisation has been used so exten- and shear stress on the sediment (Box sively to drain water from the wet soils 4.1). Moreover, water velocity and the along the streams and thereby increase characteristic linear dimension of water the cultivation of fi elds. movement can be used to estimate In all, the four parameters, slope, the Reynolds number, distinguishing discharge, plant biomass and mean- chaotic, turbulent fl ow from regular, dering, could account for 75% of the laminar fl ow. observed annual variation in mean The Danish botanist Niels Thyssen velocity in eight lowland streams. The has developed a simple empirical empirical equation is not always very model to predict mean water velocity precise in its predictions of the actual in small lowland streams (Box 4.1). velocity, however, because of restric- The velocity increases with slope, tions in its ability to account for other which generates the downstream water factors that also affect the hydraulic movement due to gravity. The veloc- resistance e.g. channel morphology, ity also increases with higher water substrate type and plant distribution. discharge because this is accompanied Mean velocity is a simple and by an increase in width and depth (and informative measure of macrofl ow. hydraulic radius) so that the resist- Danish lowland streams range from 56 RUNNING WATERS

Box 4.1 Macroﬂ ow in streams

η Mean water velocity (U, m/s) can be used to characterise Shear stress ( o) at the sediment surface is important for the stream reaches. Other suitable descriptors are depth (D, m), ability to erode particles or alternatively leave them in place. 2 width (B, m), cross section area (A, m ), slope (S, m/m) and Shear stress can be estimated from shear velocity (U*) derived meandering (I, no dimension) which can be measured as the from the logarithmic velocity proﬁ le (Box 4.2) or estimated length of the main ﬂ ow line (the thalweg) divided by the from the equation [4, 5]: shortest distance.

U* ~ 0.17 U / log (12 D/k) In small Jutland streams, the Danish botanist Niels Thyssen [1] has found the best way to empirically predict the mean water where k is sediment particle size or sediment roughness. velocity as: Shear stress in uniform ﬂ ow is given by [6]:

0.47 -0.058 0.69 -0.79 η ρ 2 U = 79.3 Q B S I o = U*

where Q is water discharge (m3/s), B is plant biomass (g dry where ρ is mass density of water (approx. 1.000 kg/m3). mass/m2), and I is an index for meandering: 1 at low, 2 at intermediate and 3 at high meandering. A direct way of measuring shear stress is to place calibrated metal hemisphere of the same size but variable weight on a The Reynolds number is the dimensionless ratio between small horizontal plate on the sediment surface and deter- inertial forces of moving water generating chaos and turbu- mine the lightest of the hemisphere set in motion [7]. lence and viscous forces attracting water parcels to move in a laminar and orderly manner, in only one direction at constant Manning’s equation is used to calculate the Manning number speed [2, 3]. If the kinematic viscosity is ν (approx. 1×10-6 (M) or the roughness coeffi cient (n = 1/M) for water move- m2/s at 20°C), the Reynolds number (Re) of main fl ow in the ment along a stream reach: stream channel is: U = 1 / nR0.67 S0.5 Re = U D / ν Where R is the hydraulic radius (surface area divided by the The fl ow is usually categorised as laminar at Re < 500, turbu- length of contact between water and sediment in the cross lent at Re > 1,000–10,000 and a mixture in between these. Re section). The coeffi cients are not constants, but empirical is already 5,000 at a velocity of 0.05 m/s and a depth of 0.10 averages of a large number of measurements. Thyssen’s m so the main fl ow in streams is almost always turbulent, comprehensive studies showed a positive relationship of although of variable intensity. In microhabitats, in sediments Manning’s n to plant biomass, but the relationship varied and within plant patches or close to the banks in still-fl owing substantially from one stream to another, preventing the streams, laminar fl ow may be encountered. construction of a general model.

reaches with rapid currents close to rents water masses pass slowly along streams, even if oxygen production and 1 m/s and a stream bed dominated the reaches, and turbulence and oxygen oxygen consumption on an area basis by stones and gravel, to reaches with exchange with the atmosphere are in the two stream types are identical. slow currents of less than 0.10 m/s and restricted. As a consequence, oxygen Together, experts and the general muddy substrata. Velocity is important concentrations in the water can change public can share the fascination of a for organisms freely exposed to the markedly between day and night variable fl ow pattern along meander- current, as drag forces increase steeply in slow-fl owing streams, while diel ing reaches with alternating pools with the water velocity. In slow cur- amplitudes are modest in fast-fl owing and riffl e sections. Every angler, who 4 Water fl ow at all scales 57

Box 4.2 Velocity proﬁ les

Figure 4.1 Main A regular velocity profi le at depth is formed in large Max. line of flow lines of fl ow at the streams with a uniform slope and cross section and a fi ne- surface of a meande- grained stream bed [8, 9]. The layer closest to the bed (the ring stream, as seen bed layer) is usually only a few millimetres thick. In slow from above. The two fl ow over a smooth bed a viscous sublayer with laminar Cross section cross sections show fl ow is formed, in which the microscopic water layers Deposition the weaker fl ow move orderly on top of each other with no vertical mixing across the stream [8]. and a linearly increasing velocity with increasing distance from the sediment (Figure 4.2). In the laminar sublayer, vertical transport is solely by diffusion. In a rough turbulent regime, a logarithmic layer is found further away

Cross section from the sediment, in which water velocity (Uz) increases linearly with the logarithmic distance (z) from the sediment:

κ Uz = U* / loge (z/zo)

κ has tried his luck with a worm and ﬂ oat, knows that the where U* is shear velocity (Box 4.1), is von Kárman’s concurrent declines close to the banks, and that it can move stant (approx. 0.40) and zo is roughness height which can upstream in the shelter behind meanders and plant patches. be estimated iteratively by linear regression of Uz versus

The angler uses his experience to attract fi sh in a territory loge z. zo can also be approximated as 1/30 of the particle chosen because of its favourable current and opportunities size for sand and 1/15 for well-sorted gravel. When shear of shelter. velocity is known, shear forces on particles and organisms A meandering stream has a certain transversal fl ow in on sediment or plant surfaces can be calculated (Box 4.1). addition to the main downstream fl ow [8]. At the apex of meanders, maximum fl ow has moved close to the outer In deep streams there is a poorly defi ned outer layer bank generating scour and bank recession. At the outer edge above the logarithmic layer. The popular assumption that of the turn, where the fl ow hits the bank, the water level is the mean velocity of the depth profi le can be measured at slightly elevated generating a secondary fl ow towards the 0.6 times of the depth requires that the logarithmic layer bottom and across the stream. At the inner edge of the turn, extends to the surface. This is often not the case. In small, water depth is reduced and the current is diminished and plant-rich streams, the logarithmic layer is not developed directed towards the water surface leading to sedimentation because of frequent acceleration and deceleration of the of particles (Figure 4.1). This transversal, spiralling fl ow can fl ow due to the presence of large roughness elements even appear in straight channels [9]. The phenomenon infl u- (stones and plant patches) and a strong and variable ences erosion and sedimentation and is, therefore, important bank roughness. In fast fl ow above an irregular sediment for the development of channel morphology and the distri- surface, the laminar sublayer is often absent, because bution of bottom dwelling invertebrates. roughness elements exceed the theoretical thickness of the laminar sublayer (δ) under ideal fl ow: Water stage is a practical parameter δ ν Water stage (depth), width, discharge and discharge capac- = 11.5 / U* ity are important stream parameters. Water depth and width set the limit for the existence of large fi sh and plants, and Like the logarithmic layer, the laminar sublayer is never deep water increases the diversity of small invertebrates well defi ned (e.g. the constant used can vary between [10]. Water depth is also important for many physical proc- 8 and 20). Overall, the hydraulic equations should be esses. Greater water depth reduces the physical contact viewed as guide lines and not as physical laws. between the water volume and the atmosphere, thereby 58 RUNNING WATERS

restricting the exchange of oxygen, as the discharge that can pass the cross A carbon dioxide and heat. Greater water section at a given water stage, and is depth also reduces the contact between thus a measure of the water volume Outer layer the water volume and the stream bed that can be drained per unit time from and thereby the frictional resistance to a given area without the water level fl ow. Finally, water depth infl uences reaching the mouth of the drain pipes Logarithmic layer the velocity profi le (Box 4.2, Figure or the upper level of the banks causing 4.2) as greater depth combined with fl ooding of the fi elds. Height above stream bed Height above stream unchanged mean velocity reduces Discharge capacity changes over Laminar sub-layer friction velocity and shear stress at the the year. It is usually highest in late Velocity bottom permitting muddy bottoms to winter when plants are absent and fi ne- form, despite a relatively high mean grained material has been transported 1 B velocity (Box 4.1). away by high fl ows. In this situation

(m) Water stage has been measured the Q-H relationship follows a “basic d log z continuously at numerous stations curve” (Figure 4.3). During spring and 0.1 in Danish streams for more than 70 summer there is a gradual curve shift d U z years. These measurements are often of the relationship, concurrent with d Uz/d log z μ U 0.01 * combined with monthly or biweekly increasing plant biomass and the accu- measurements of water discharge mulation of bottom material. Discharge Zo deﬁ ned as the water volume passing capacity can therefore be described

Height above stream bed Height above stream 0.001 a given cross section per unit time by a set of curve variants based on 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 (unit: m3/s). The relationships between actual measurements. It is possible to

Velocity Uz (m/s) discharge (Q) and water stage (H) at a calculate a theoretical curve of water given site can be described by power stage (showing “the reference stage”) at b Figure 4.2 Vertical depth proﬁ le of ideal functions of the type: Q = a (H – Ho) a selected constant discharge (i.e. “the water velocity. which form straight lines by logarith- reference discharge”). Often the mean A: A thin viscous sub-layer with laminar mic transformation (i.e. log Q = log a + discharge for the summer or the entire

fl ow is found above the stream bed (max. b log(H – Ho)). Ho is the water stage at year is used as the reference for evalu- 2–3 mm), followed by a logarithmic layer a discharge of zero, while “a” among ation. The reference water stage will and an outer layer with velocities approa- other things refl ects the resistance to reveal the infl uence of accumulated ching a constant level. fl ow, and “b” depends on channel plant biomass and sediment during B: In the logarithmic layer, velocity increases morphology. Discharge capacity is of spring and summer and the later linearly with the logarithm to distance from great interest to farmers. It is defi ned reductions in autumn and winter. κ the sediment with a slope of /U* There are different ways of calcu-

(approx. 0.4/U*). Redrawn from [9]. lating the water stage at a selected discharge at a given time and plant Figure 4.3 The water stage in streams biomass. The “proportionality method” High biomass increases with water discharge. When the assumes that water stage increases at discharge is the same/constant, water stage the same percentage relative to the Low biomass is lower in winter than in spring. Plants and “basic curve” at all discharges [11]. Basic line ﬁ ne sediments are usually washed out in This method is applicable in streams

Water stage Water winter, and the discharge-stage relationship with few plants and within narrow follows the “basic curve”. With the same ranges of water discharge. The “focal discharge, the water stage is higher in point method” assumes that all Q-H summer because of ﬂ ow retardation due to curve variants intersect in a common Water flow plants and accumulated sediments. focal point at high discharge. The 4 Water ﬂ ow at all scales 59

assumption is that ﬂ ow retardation Figure 4.4 The Broad-leaved Pondweed 1 caused by the plants is relatively more drag on plants in important at low water stage and ﬂ owing water incre- = 1.57 = 0.58 – – declines at higher stages. The focal ases with increasing B = 19.4 g U = 0.40 m/s 0.1 point method is commonly used and current velocity and is the most realistic method for very plant biomass.

(Newton) = 2.17 – α: slope of the log- B = 1.4 g variable discharges [12]. It may be less 0.01 suitable when the vegetation consists log relationship, Drag = 0.67 – predominantly of stiff species (e.g. U: mean velocity, U = 0.15 m/s amphibious and emergent species close B: mean biomass. 0.001 to the banks), which do not bend over • = 4 shoots, o = 64 and become ﬂ ooded at peak discharge shoots. Fennel Pondweed 1 in the same way as the ﬂ exible sub- = 1.56 = 0.78 – – merged plants. B = 48.7 g U = 0.40 m/s

0.1 Water level and plants = 1.63 The inﬂ uence of plants on water –

(Newton) B = 5.5 g level has been evaluated in 65 Danish 0.01 = 0.86 streams [11]. The presence of plants Drag – during summer meant an average U = 0.15 m/s increase of 7.5 cm in the water level, 0.001 for a constant mean summer discharge, 0.1 0.5 1 0.1 1 10 100 and in a few cases the increase was as Current velocity (m/s) Biomass (g) high as 25–30 cm. The effect is most evident in mid-summer when the actual discharge is low. During high discharges and strong currents, the they grow scattered across the stream also offer better shelter to downstream submerged plants bend over and are bed and each shoot is freely exposed neighbours than fl exible, streamlined pressed towards the stream bed, and to the fl ow. When they are distributed shoots. As a consequence, the hydrau- a large proportion of the water passes in dense patches, mutual fl ow protec- lic resistance does not increase propor- over the plants unimpeded. Amphibi- tion is possible among the shoots, and tionally to the plant biomass as shoot ous and emergent plants in shallow much of the water can pass unimpeded density increases. Experiments in labo- water and on the banks are relatively in the fl ow channels between the plant ratory fl umes show that fl ow resistance stiff and remain resistant at increasing patches. of the plants often increases 4 to 7 times discharges and water levels. Plant species have a variable resist- as the biomass increases 10-fold (Figure The retardation of fl ow is highest ance to fl ow when they grow at low 4.4). Moreover, the resistance of differ- along reaches with a dense plant cover densities. Single shoots of stiff, ramify- ent species tends to converge towards and a modest downstream slope. No ing species are more resistant to fl ow a common value at high plant biomass fi eld measurements have addressed the than fl exible, streamlined species. These and high water velocity. The plausible effect of the different architectures of differences between species are less explanation for this behaviour is that plant species and their spatial distri- apparent at high shoot density within species having a high fl ow resistance as bution on fl ow resistance. However, plant patches, because of mutual pro- single shoots exposed to low veloc- fl ume experiments and theoretical con- tection and shelter among the shoots. ity will provide better shelter among siderations lead to the suggestion that Stiff, ramifying shoots exposed to full- neighbouring shoots when the biomass at a given biomass, submerged plants scale fl ow at the patch margin experi- and velocity increase than species offer- increase the water level most when ence a substantial drag but these shoots ing less resistance as single shoots. 60 RUNNING WATERS

Winter reduced [13]. In several cases, however, 15 it was difﬁ cult to establish unequivo- cally the reasons for an altered water 10 level.

5 Plant patches 0 Stream plants commonly form a Summer mosaic of individual plant patches 15 across the stream bed. From year to year plant patches change position and 10 different species may dominate. Plant

5 patches diminish in size or disappear during the winter, after cutting or

Number of streams Number of streams 0 episodes of peak fl ow. They recover < –20 –10 0 10 20 >30 from surviving roots and rhizomes Change of water level (cm) buried in the stream bed, or grow from seeds and drifting shoots being Figure 4.5 Directional change in mean trapped by stones and other obstacles Cushions of Zannichellia palustris (horned water level during winter and summer in on the sediment surfaces [14]. A single pondweed) in the Vindinge Stream. 65 Danish streams from 1976 to 1995. More shoot is capable of forming a patch streams have experienced an increase in of several square metres during the water level than a decline [11]. course of a few summer months. The daily linear expansion of patches is, for fast fl ow and erosion become concen- example, typically 0.3–2.0 cm for Cal- trated in the fl ow channels between the litriche (water starwort) and 0.5–4.0 cm patches. Water stage and management for Ranunculus (water crow-foot) and Individual patches tend to behave The introduction of more gentle man- Sparganium (bur-reed). as functional units. A single shoot often agement of large Danish streams, in The disturbed environment together forms a large plant patch with a shape the 1980s and 1990s, had the objective with the effective dispersal and the modulated by the current, while the of reducing anthropogenic distur- vegetative expansion of stream plants patch generates its own characteris- bance, diminishing cross section and all contribute to forming the scattered tic fl ow and sedimentation pattern increasing water level. In a study of 65 mosaic of single-species patches. Cal- [15]. Studies of Callitriche have shown streams, 32 experienced a markedly litriche, Elodea canadensis (Canadian that patches become fl atter and more higher water level for a given dis- waterweed) and Ranunculus form streamlined with greater impact of the charge, while 28 remained unchanged, dense patches with a distinct margin, current. Viewed from above, the green and 5 experienced a marked decline in while Sparganium have long horizon- canopy and the rooted area of exposed water level (Figure 4.5). tal runners and form open patches. patches resemble an ellipse (Figure 4.6). Streams developing a markedly Reaches with unconsolidated sand A side-view shows the canopy as nicely higher water level have in several cases sediments have a greater tendency to rounded upstream and the height is experienced a more gentle manage- develop a patch mosaic than reaches often only 10–20% of the length in ment regime. Increased water levels with stable bottoms, the reason being order to minimise form drag. were for example observed in streams that uprooting of new colonising The elongated, streamlined shape is experiencing increased sedimentation shoots is more likely outside the estab- mainly observed in shallow (0.1–0.3 m) and consolidation of the stream bed in lished patches in reaches with unstable riffl es where the fl ow is accelerated the near-shore zone where the inten- sediments. Plant patches stabilise the along the sides of the patches, and the sity of cutting and dredging had been vegetated sediments underneath, while shoots are pressed against the sedi- 4 Water fl ow at all scales 61

ment. The elongated shape is a result the shape of the canopy and the char- of daily growth rates of, for example, acter of water movement around the 1.5 cm in downstream direction in the entire patch margin and the individual shelter of mother shoots. Upstream shoot surfaces [2, 3].

growth rates against the fl ow are often The coeffi cient of friction drag (Cf close to zero and lateral daily growth in Box 4.3) declines with increasing rates from both sides may be 0.3 cm. velocity so that friction drag increases With these growth rates a single shoot less rapidly than with the square of will form a 1.5-m long and 0.6-m wide velocity; often in approximate direct patch in 100 days. In deeper reaches of proportion to the velocity. At increas- slower velocity, fl ow can more easily ing velocity the individual shoots and pass above the patches which become the entire canopy may become more higher and approach a circular shape compressed against the stream bed, because expansion can take place at the thereby reducing the frontal area. same rate in all directions. Turbulence behind the plant patch may also change, but the width of the turbu- Flow and drag in plant patches lent band and the pattern of turbulence Flowering bur-reed. The current generates drag in the green (e.g. size and spatial distribution of canopy of plant patches, while the vortices) are diffi cult to predict. The roots keep these anchored. Shoots and net result is that form drag increases roots must therefore develop a mutual at higher velocity with an exponent balance between one another depend- between 1.0 and 1.75 (i.e. a 10–56 times A ing on the fl ow regime and the strength higher drag for a 10-fold higher veloc- of anchoring. ity). In rare cases the exponent can There are three types of drag forces exceed 2.0 (Figure 4.4). Because form in streams (Box 4.3). Form drag is drag is caused by pressure phenomena, generated by pressure loss from the while friction drag is the result of vis- upstream exposed front of the canopy cous forces, form drag will increase in to the downstream part behind the importance relative to friction drag at canopy. Friction drag is due to fric- increasing velocity and increasing size tion between the moving water and of organisms (i.e. at a higher Reynolds plant surfaces. Lift is generated by the number). Because plants have a large B acceleration of water above the canopy surface area per plant biomass in order surface. Form and friction drag act to absorb sunlight for photosynthesis, in the direction of fl ow, while lift acts friction drag is never insignifi cant. Figure 4.6 The most common shape of 50 upwards perpendicularly to the fl ow. In very dense plant patches of Cal- plant patches of Callitriche platycarpa (vari- All three forces rise in proportion litriche, much of the fl owing water is ous-leaved water starwort) found in small to the mass density of water and the diverted above and along the sides of Jutland streams in the summer. The canopy square of velocity. Form drag increases the canopy, while the interior remains is almost elliptical when viewed from above in proportion to the frontal area facing sheltered (Figure 4.7). As a result, form (A). From the side (B) the canopy appears the current, friction drag increases in drag and lift are high for the entire fl at and rounded upstream with a long proportion to the plant surface and lift canopy, but forces are concentrated unattached canopy trailing downstream. The to the projected area (Box 4.3). All these at the exposed shoots at the canopy green canopy covers a much larger area than parameters are measurable and can be surface, while the shoots in the interior the roots. Arrows indicate the direction and evaluated. Each force is infl uenced by experience reduced forces. In contrast, velocity of the fl ow [Sand-Jensen, original]. an odd coeffi cient which depends on the fl ow passes through the open cano- 62 RUNNING WATERS

Box 4.3 Drag forces

Flowing water generates three important dynamic forces: pies of Sparganium at relatively high velocities (Figure 4.7) so

form drag (Fp), friction drag (Ff) and lift (FL). The unit is that form drag and lift for the patch as a whole are low, while Newton (kg m/s2) and the equations are: form drag is higher and more equally distributed among the individual shoots. It appears to be a suitable adaptation for ρ 2 Fp = 0.5 U Af Cp Callitriche to have fragile, ﬁ nely dissected shoots and stream- ρ 2 Ff = 0.5 U Aw Cf lined canopies, and for Sparganium to have strap-formed, ρ 2 FL = 0.5 U Ap CL streamlined leaves and patches of a less streamlined shape. As plant patches grow in size there is a parallel increase ρ where is mass density of water, U is water velocity, Af in the frontal area facing the current, the surface area viewed

is frontal area of the object facing the current, Aw is the from above and the total surface area or volume. Studies on

entire surface area of the object and Ap is projected area fl ow-exposed patches of Callitriche show that the sediment perpendicular to the fl ow. The three dimensionless coeffi - area covered by roots increases predictably with the size of

cients are: Cp for form drag (1.0 for a ﬂ at plate facing the canopy. On average, the frontal area increases sixfold and current, approx. 0.01 for the edge of the plate facing the the total area 14-fold when the rooted area increases 10-fold

current, and < 0.2 for streamlined objects), Cf for friction (Figure 4.8). It is appropriate that the frontal area increases drag (declining at increasing water velocity and size of less than the rooted area, because larger patches tend to

objects) and CL for lift. Coeffi cients for certain geometric grow taller and are therefore exposed to higher velocities forms and organisms are found in [2, 3]. The coeffi cients further away from the sediment than are small patches with should not be regarded as constants because they change shoots tightly compressed against the sediment. That the with the character of the fl ow. development of shoots and roots are in fact well adjusted appears from the fact that patches located in the shelter close Form drag and friction drag act in the direction of the to banks and behind turns develop more shoots as compared fl ow, while lift acts perpendicular to the fl ow. A fourth to roots than patches freely exposed to the current in the force, the acceleration force, is usually insignifi cant in middle of the stream. streams because of small accelerations, while it is important for large objects on wave-swept shores in the sea and Flow and the individual in lakes [2]. The infl uence of fl ow on individuals is two-sided. On the positive side, fl ow continuously supplies new, fresh water and reduces the laminar sublayer surrounding the surfaces of organisms and, thereby promoting the exchange of gases and solutes with the surrounding water [18]. At low

concentration levels of dissolved nutrients, CO2 and oxygen in the water, the biological uptake increases at gradually higher concentrations until saturation is achieved. This point of saturated uptake is reached at gradually lower external concentrations as the water velocity increases (Figure 4.9). The response is logical, because the biological uptake rate increases in proportion to the steepness of the gas and solute gradients across the laminar sublayer. Great differences in concentration between the free water and the surface of organisms and the thinner laminar sublayer will both lead to steeper gradients and promote the uptake. Thus the risk of suffocation of fi sh and invertebrates by oxygen depletion is highest when low oxygen concentrations and low water Small stream with localised patches of fl owering Ranunculus. velocities coincide. 4 Water fl ow at all scales 63

Starwort Current velocity (cm/s) 0 0.250.50 0.75 0 0.250.50 0.75 0 0.250.50 0.75 0 0.250.50 0.75 0 0.250.50 0.75 0 0.250.50 0.75 0 I II III IV V VI –10

(cm) –20

–30 Depth

–40

–50 0 0.25 0.50 0 0.25 0.50 0 0.25 0.50 0 0.25 0.50 0 0.25 0.50 0 0.25 0.50

Relative turbulence (% of mean velocity) Upper edge of canopy

Unbranded bur-reed Depth profiles I II III IV V IV Current velocity (cm/s) 0 0.250.50 0.75 0 0.250.50 0.75 0 0.250.50 0.75 0 0.250.50 0.75 0 0.250.50 0.75 0 0.250.50 0.75 0 I II III IV V VI –10

(cm) –20

–30 Depth

–40

–50 0 0.25 0.50 0 0.25 0.50 0 0.25 0.50 0 0.25 0.50 0 0.25 0.50 0 0.25 0.50

Relative turbulence (% of mean velocity)

Figure 4.7 Current velocity and relative On the negative side, fl ow gener- to invertebrates becomes critically low, turbulence in patches of Callitriche copho- ates a physical disturbance, which can the invertebrates can select the most carpa (starwort) and Sparganium emersum induce metabolic stress and a greater elevated and exposed parts of the (unbranched bur-reed) were measured in risk of being transported downstream stream bed with the highest velocity. depth profi les along the mid-axis from with the current or being crushed or Some species can actively enhance ven- upstream (I) to downstream of the patches buried below moving sediment parti- tilation by moving the gills or the abdo- (VI). Arrows mark the upper edge of the cles. The metabolic stress may be costly men or by moving legs and toes rapidly canopy. Redrawn from [16, 17]. due to energy investment in greater to push the body up and down [4]. basal respiration. The fl apping movements of leaves In order to balance the advantages and stems of submerged plants will and disadvantages of the fl ow, spe- enhance respiration. In response to cies have evolved adaptations that these movements, the plants tend to allow adjustment of their morphology, grow more slowly, develop smaller anatomy, habitat selection and behav- cells with thicker walls, and form more iour. For example, when oxygen supply robust shoots with shorter, denser and 64 RUNNING WATERS

Adaptation to rapid ﬂ ow and 2 A disturbance )

2 Plants and animals adapt to high 0.2

(m velocity by reducing or resisting the impact. Common adaptations include: 0.02 1) streamlined shape, 2) small fron-

Front area Front tal area, 3) compression against the 0.002 substratum, 4) large anchoring surface, and 5) high anchoring strength. Not 10

) B 2 just one, but several possible adapta-

(m tions and combinations involving dif- 1 ferent advantages and disadvantages are observed. 0.1 In the case of the plants, a compact Canopy area form and compression against the 0.01 substratum reduce drag forces, but also 1 the exposure to the surrounding water C and to incoming light makes it harder Beautifully sculptured patches of starwort. ) 3 0.1 to obtain the necessary resources for (m growth. If the canopy is instead freely 0.01 exposed to the water and composed Volume of thin tissue, it is easier to sequester 0.001 the necessary resources, but the risk of streams with slow fl ow. Blackfl ies have 0.01 0.1 1 excessive drag will require streamlin- exceptionally strong anchorage because Root area (m2) ing, effi cient anchoring and resistance they spin a net to the surface and attach to breakage of the shoots. Ranunculus , to it by means of hooks on the base of Figure 4.8 Relationship between size of for example, grow freely exposed to their abdomen [4]. the green canopy and the surface area of rapid currents which are tolerated A traditional belief among stream root anchoring of 50 patches of Callitriche because of streamlined, capillary leaf biologists has been that the compressed cophocarpa in small Jutland streams during fi laments, strong, fl exible stems and body shape of invertebrates allowed summer. Front area facing the current (A), long roots that wind around stones these to live safely and sheltered within canopy area as seen from above (B) and and cobbles. In shallow water with the laminar sublayer immediately volume occupied by the canopy (C) [Sand- very rapid fl ow, the Ranunculus shoots above the substratum (Box 4.4). How- Jensen, original]. become shorter to reduce drag, and the ever, the animals are often much higher patches become anchored along most (2–5 mm) than the expected thickness of their length. of the laminar sublayer (e.g. 0.1–0.5 Many stream invertebrates are fl at mm) and in rough turbulent fl ow the stiffer leaves. However, the most recent and able to adhere closely to the sub- laminar sublayer does not exist at all studies suggest that stream plants are stratum and they have strong legs and and animals are exposed to high veloci- capable of maintaining growth at rela- claws. Several species have character- ties and turbulence. The animals can tively high velocities, probably because istic positions or postures that reduce cling to and press themselves against the fl ow is relatively predictable in drag and ensure down force. Among the substratum and thereby control and both direction and velocity. If the plants some species of snails, the “foot” resist the form drag and lift caused by are exposed to a very variable fl ow grows relatively large and the anchor- the current, but they cannot eliminate regime (e.g. a short daily peak veloc- ing strength in rapid fl ow increases or avoid the forces at exposed solid ity), growth is further inhibited. as compared with species living in surfaces. 4 Water fl ow at all scales 65

Box 4.4 Turbulent or laminar ﬂ ow

Despite the adaptations, the organ- It was previously believed that benthic invertebrates always experienced laminar isms experience strong disturbances fl ow. This is probably the case when they live deeply buried in sediments or in during peak fl ow. The risk of mortal- non-ventilated burrows. Exposed at the sediment surface, however, they experi- ity is greatest when the stream bed is ence signifi cant velocities and turbulence. Turbulent fl ow means that fl ow at a eroded and set in motion. The plant given point changes direction and velocity with time due to passage of vortices. and animal life in streams therefore Turbulence allows a greater supply rate of oxygen to animals, while the draw- alternate between episodes of substan- back is a greater physical impact and risk of drift with the current. tial loss and longer, calmer periods of recolonisation and reestablishment. Flow was earlier believed to be laminar through plant patches. Water velocities The dynamic conditions for attached within dense, free-standing plant patches often decline to 10–20% of the unre- microalgae, whose biomass declines stricted velocity outside the patches [17]. The smallest velocities were about 2 steeply during storm fl ow events and cm/s within canopies of Callitriche cophocarpa. At a characteristic linear dimen- increases gradually during periods of sion of 1 cm for the fl ow within the canopy (e.g. the size of Callitriche leaves low disturbance and suitable growth or the distance between the leaves), the estimated Reynolds number is 200. conditions, have been most extensively Nonetheless, the fl ow remained turbulent at an intensity of more than 0.2 cm/s documented. (standard deviation of velocity). The intensity of turbulence increases linearly Documentation from New Zealand with mean velocity both in the free water and between the plants, but turbu- streams shows that species richness lence intensity relative to mean velocity was enhanced 1–4 cm above the sedi- of benthic invertebrates depends on ment and canopy surfaces where fl ow is strongly retarded (Figure 4.7). Pressure the frequency and extent of physical differences and leaf movements are regarded as important for maintaining disturbance generated by peak fl ow turbulence, despite low velocities in the cm-sized linear dimensions within cano- [19]. The studies support the inter- pies. It is likely that water velocity and turbulence are suffi ciently reduced within mediate disturbance theory predict- dense plant patches to permit the development of laminar sublayers immedi- ing that species richness is highest at ately above plant surfaces. intermediate disturbance. At very high disturbance only a few very robust species survive, and at low disturbance the competitively superior species have the suffi cient time to out-compete the Figure 4.9 Metabolism of aquatic orga- weaker species during the extended nisms commonly increases at higher exter- Figure 4.10 Species richness of invertebra- stable periods (Figure 4.10). The species nal concentration and current velocity until tes in New Zealand streams as a function richness of aquatic plants in English saturation is reached. The ideal relationship of the proportional area of the stream bed canals also appears to exhibit the maxi- applies to photosynthesis, growth and set in motion at peak velocities. Maximum mum species richness at intermediate uptake of nutrients and inorganic carbon of species diversity is observed at medium disturbance, although differences are plants and respiration of animals. disturbance. Redrawn from [19]. relatively small between sites of low and intermediate disturbance [20]. Strong current 35 Because intermediate water veloci- 30 Average ties are optimal for metabolism and 25 current intermediate disturbances are optimal 20 for species richness, it appears that Weak current

a compromise is the ideal situation Nutrient uptake 15 Number of species here as it often is in nature and life in 10 general. 8 10 20 30 4050 100 Nutrient concentration Disturbance (%) 66 RUNNING WATERS

Plants form a gradual transition between land and water. 67

5 Aquatic plants Aquatic ﬂ owering plants form a relatively young plant group on an evolutionary timescale. The group has developed over the past 80 million years from terrestrial ﬂ owering plants that re-colonised the aquatic environment after 60–100 million years on land. The exchange of species between terrestrial and aquatic environments continues today and is very intensive along stream banks. In this chapter we describe the physical and chemical barriers to the exchange of plants between land and water.

Broad-scale patterns fl owering and pollination. Other spe- Unique characteristics of streams and cies have runners that grow from land rivers are the unidirectional fl ow of to water to establish submerged popu- water and the intimate contact with lations. And yet others are essentially the surrounding terrestrial ecosystems. terrestrial plants that grow submerged Assuming an average width of Danish when the water table is high. streams of one metre, the riparian con- The fact that plant species can move tact zone is an estimated 130,000 km from one environment to another, or close to 3 km bank per km2 of land. in this case between land and water, Across this transition zone between makes it diffi cult to quantify the land and water is an intense fl ow of number of genuine water plants. The water with its content of dissolved Danish fl ora contains about 1,265 spe- and particulate inorganic and organic cies of fl owering plants [1]. Of these, matter. about 50 species are truly or obligate The dense animal and plant popula- submerged plants with permanently tions in the riparian zone allow an submerged shoots, 55 species are active exchange of species between amphibious species with the ability to the two environments. A well-known grow submerged as well as emerged example is the dipper – a bird that for- and 75 species are terrestrial plants that ages on invertebrates under water, but can survive submergence for shorter or otherwise lives on land. Many insects longer periods (months), but can never live in water during the juvenile stages, form permanently submerged popula- whereas the adults live on land, only tions or develop distinct water forms to return to water to lay their eggs. (Table 5.1). Among plants, many terrestrial-aquatic If we expand our time span and transitions also exist. Some species live consider the aquatic fl ora from an evo- submerged, but raise their fl ower stalks lutionary perspective, our impression up above the water surface during of the riparian zone as an area with 68 RUNNING WATERS

Table 5.1 Number Number of species of species of the Primary aquatic plants 51 various plant Amphibious plants 55 growth forms in Temporarily submerged 75 the riparian and terrestrial plants submerged zone Danish plants – in total 1,265 of Danish streams.

intense fl ow of plant species from one environment to the other is confi rmed. Terrestrial plants are believed to have evolved from a group of freshwater green algae that colonised terrestrial habitats 300–400 million years ago and diversifi ed under variable terrestrial conditions. One result of this diversifi - Berula erecta is an cation and evolutionary development amphibious plant was the fl owering plants. During the that looks the same past 80 million years several of these on land as under plants have recolonised aquatic habi- water. tats, and their descendants constitute the majority of the plant species currently inhabiting streams and lakes. Irrespective of the time span, colonisation of aquatic habitats requires adaptation by the plants to the physical, chemical and biological conditions prevailing in the aquatic environment. In this respect, it is a general belief that water is an unfavourable environment for plant growth, and that this is the reason for the rather low species richness observed in freshwater habitats. We believe, however, that this view is unbalanced, as the plants will have both advantages and disadvantages from growing submerged, in the same way as they have advantages and disadvantages from growing on Aquatic Ranunculus land. Thus, while some environmental species produce factors might constitute a barrier to beautiful white the colonisation of a new environ- fl owers in air above ment, other parameters may be more the water where benefi cial to plant growth. In line with they are pollinated this view, it should be noted that the by insects. 5 Aquatic plants 69

The other type comprises factors that affect plant growth more indirectly by inﬂ uencing plant metabolism without acting as resources. Among the indirect factors are temperature and other physical and chemical parameters (e.g. oxygen, pH, phytotoxins). For stream plants in particular, the moving water is of great importance for growth and distribution. The force of the moving water can be vigorous and may up-root or prevent establishment of plants on the stream bottom. Establishment is prevented either because the plants are broken by the force of the water or because the rough substratum with gravel and stones, which is a result of a high ﬂ ow velocity, is unsuitable for rooted plants. These aspects are discussed in more details in other chapters and will not be considered here. We will focus on resource limitation and acquisition.

Light and minerals Light availability is restricted in aquatic habitats as compared to terrestrial systems due to refl ection from the water surface and a high light attenuation by water with its content Several terrestrial of dissolved and suspended material. Despite this loss, how- plants can sustain ever, light is hardly a signifi cant barrier to plant colonisation species richness in freshwater habitats short periods of of aquatic habitats. First, aquatic plants grow close to the is not particularly low, considering that submersion. surface where irradiance is reduced by only a few percent terrestrial habitats cover an area that is (10–15%) relative to irradiance above the water. Growth is about 50 times larger than that of the only prevented if the irradiance above the water is low, as in aquatic habitats, and that only a small forests, or the light is restricted due to intense shading from fraction of aquatic habitats is actually the banks. Second, acclimation and adaptation to low irradi- populated by plants. ance is a common phenomenon among terrestrial plants We will consider the various factors and consequently among the predecessors of aquatic plants. to which plants have to adapt, and to A good example is the herb and understory vegetation in assess the importance of barriers for forests. the colonisation of aquatic habitats Terrestrial plants adapt to low irradiance by reducing by terrestrial plants, that might be of biomass allocation to non-photosynthesising tissues, and importance. We defi ne a barrier as a thereby reducing respiratory costs. As a result, shade- factor when it prevents plants from adapted plants often have thin leaves, low root biomass, growing, unless they show specifi c a low root:shoot biomass ratio, and less structural tissue traits that allows growth despite the in stem and leaves than sun-plants. These strategies can negative impact of the specifi c factor. be fully exploited by aquatic plants. In fact, reduction in There are two types of important structural tissue may be more readily achievable for aquatic factors affecting colonisation and plants than terrestrial plants due to the high buoyancy in plant growth in aquatic habitats. One water. type comprises the resources needed Although light may not constitute a signifi cant barrier directly for growing, i.e. light, inor- to plant colonisation of aquatic habitats, light is still a very ganic carbon, oxygen and minerals. important factor that regulates plant growth and distribu- 70 RUNNING WATERS

Box 5.1 Nutrient supply to aquatic plants

For plants the main source of mineral nutrients is the sediment, while the incor- ment. The ample supply of nutrients in poration of nutrients in organic matter often takes place in leaves and shoots. Danish streams results from high con- As a result, the nutrients need to be transported from the roots to the leaves. In centration in both water and sediment. terrestrial plants the nutrient are carried in the transpiration stream, which is a Nutrient availability in the sediment fl ow of water from root to shoot driven by evaporation from the leaf surfaces. is further enhanced by sedimentation Evaporation “sucks” water from the hydro-soil through roots and stem to the of fi ne particles within the dense plant leaves and fi nally releases it to the atmosphere as water vapor. patches typical of most aquatic plants species growing in streams. Evaporation is not possible in submerged macrophytes. However, these plants still have a need for nutrient transport from roots to shoots. The transport can Oxygen and carbon dioxide take place by diffusion from cell to cell in short plants, but for tall plants this Light and nutrients do not appear to process is much too slow. Research shows that aquatic plants are able to build constitute signifi cant barriers to plant up a root pressure suffi cient to drive a signifi cant fl ow of water up through the colonisation of aquatic habitats. So the stem of the plants to the leaves. At the leaf rim the water is exuded through dogma that aquatic habitats are unfa- specialised pores called hydathodes. vourable to plant growth must rely on the availability of the metabolic gases It is technically diffi cult to measure the root pressure in aquatic plants. But oxygen and carbon dioxide. preliminary measurements show that pressures above 0.3 kPa are attainable. Oxygen is essential in the break- Measured fl ow of water varies among species from 0.2 to 3.5 ml/g plant per day. down of organic matter in respiration, These rates are suffi cient to cover the need for nutrients under most natural and carbon dioxide is the substrate in conditions [2]. photosynthesis and the carbon source used in synthesis of organic matter. Carbon dioxide is easily dissolved in water and when air and water are tion. Thus, in Danish streams no rooted In the aquatic environment the in equilibrium, the concentration of plants are found at water depths above water fl ow driven by transpiration is carbon dioxide in air is almost equal approx. 1.7 m due to lack of light. blocked since transpiration is absent. to that in water (approx. 16 mmol/m3 With respect to nutrient acquisition However, there is still a signifi cant at 15°C). Oxygen is less soluble and it is likely that adaptation to nutrient- transport of water driven by root pres- the concentration in water is about 30 poor or nutrient-rich conditions sure (Box 5.1). The fl ow rate is lower times lower than in air when in equilib- evolved prior to the return of higher than fl ow driven by transpiration, rium (320 mmol/m3 in water compared plants to the aquatic environment. And but calculations have shown that the to 9,400 mmol/m3 in air at 15°C). there is no reason to expect the availa- fl ow is suffi cient to meet the nutrient The solubility of gases plays only bility of nutrients to differ signifi cantly requirements by the plants. In addition, a minor role in the regulation of plant between aquatic and terrestrial envi- submerged aquatic plants have the metabolism, however, since all physi- ronments. An assessment is diffi cult, capability to acquire nutrients from the ological and biochemical processes in however, and has to include not only bulk water surrounding the leaf. the plants take place in solution in the a comparison of nutrient concentra- For rooted plants in Danish streams, cell sap. What is more important is tions in the two environments, but also studies conducted so far indicate that the 10,000 times lower diffusion rate physiological and spatial aspects. the availability of nutrients is suffi cient of gases in water compared to air. As Terrestrial plants acquire nutrients to saturate growth. This conclusion a consequence, the diffusion of gases through their roots and the nutrient is based on measurements of tissue across the layer of stagnant air or water ions are carried from roots to leaves concentrations of the most critical surrounding the leaf surfaces (the dif- by water transported in the vascular nutrients [3] and on fi eld experiments fusive boundary layer) is substantially system driven by transpiration from where the availability of nutrients was lower in water compared to air. The the leaf surfaces. enhanced by spiking water and sedi- effect is greatest at low fl ow velocities, 5 Aquatic plants 71

4 Elodea canadensis

Berula erecta 2 (%/day)

Growth rate 1

0 0 5 10 15 20 25

Flow velocity (cm/s)

Plants form dense stands in the riparian zone between land and water. Figure 5.1 Flow velocity has a marked positive infl uence on the growth rate of submerged aquatic plants. Elodea canadensis (brown curve) and Berula erecta (green where the thickness of the bound- to the leaves under natural growth con- curve). [J. Bagger, unpubl.]. ary layer is greater (Chapter 4). Thus, ditions might be of great importance as rates of photosynthesis measured in a growth-regulating factor (Figure 5.2). the laboratory on leaves or shoots of For plants in the riparian zone the 8 submerged water plants increase with low supply of carbon dioxide in water Callitriche sp. increased fl ow velocity until a maxi- might constitute a barrier to colonisa- 6 mum is reached at velocities of 1–5 tion of the aquatic environment. There- cm/s [4, 5]. At velocities above this fore, species that can ameliorate the 4 range, the rate of photosynthesis once limitation by carbon dioxide will have Elodea canadensis (%/day) again declines, presumably due to the a competitive advantage over spe- Growth rate 2 physical stress imposed by the fast- cies lacking this ability. Two acclima- moving water. The optimal fl ow veloc- tion strategies seem prevalent among 0 ity for photosynthesis is comparable to aquatic plants. One strategy diminishes 0 200 400 600 800 3 fl ow rates measured in plant patches in the importance of the boundary layer, CO2-concentration (mmol/m ) streams (approx. 0.5–3.5 cm/s) [6]. and the other allows the plants to Flow is expected to affect isolated exploit other inorganic carbon sources Figure 5.2 The growth rate of submerged leaves and shoots differently than in addition to carbon dioxide. The fi rst aquatic plants is often limited by the availa- plants growing in dense plant stands, strategy is discussed in more details bility of carbon dioxide in the water. Limita- where the complex architecture of the below, the other in the next section. tion is greater for Callitriche (only capable of plants will modify fl ow conditions The thickness of the boundary layer using carbon dioxide) than for Elodea (able close to the leaves. As a consequence, is dependent on the actual size of the to use bicarbonate in addition to carbon laboratory experiments of the type leaf and increases with the length of dioxide). Elodea canadensis (brown curve) described above cannot be extrapolated the leaf, measured in the direction of and submerged form of Callitriche sp. (blue to fi eld conditions without critical the fl ow. No systematic studies of leaf curve) [T. Madsen, unpubl.]. assessment. In line with this, growth sizes of aquatic and terrestrial plants experiments in the fi eld using intact have been performed. However, it is shoots have shown an increase in generally believed that submerged growth rate with fl ow velocities up to plant species have small or dissected 20 cm/s (Figure 5.1). This result indi- leaves. It is also a general trait that the cates that the supply of carbon dioxide leaves of submerged plants are very 72 RUNNING WATERS

rarily submerged terrestrial plants. They will benefi t, of course, when submerged, but will suffer when they become emerged during periods with a low water table. And since amphibious Rorippa amphibia plants often grow close to the bank, it is (Great Yellow-cress) most likely that they, in an unpredict- is an amphibious able manner, are submerged during plant that grows some periods and emerged in others. along the banks When the plants become emerged they of slow-fl owing will suffer from extensive water loss streams and rivers. due to evaporation and risk dehydra- tion and wilting if the cuticle is highly permeable to gases. thin, often only two to six cell-layers thick (Figure 5.3). Leaf thickness has only a very limited effect on the thickness of Bicarbonate use the boundary layer, but the lower number of cells per unit All obligate submerged plant species, leaf-surface area reduces the number of cells that have to be with the exception of the bryophytes, supplied with carbon dioxide through that surface. have the ability to use bicarbonate in The cuticle, a continuous waxy fi lm on the external addition to carbon dioxide. The ability surface of the leaves, is thin and is expected to be highly to use bicarbonate gives the plants permeable in aquatic plants. The cuticle protects terrestrial access to an inorganic carbon pool plants from water loss. This advantage might turn out to that is up to 100 times larger than the be a disadvantage to submerged plants, because the low carbon dioxide pool (Box 5.2). How- permeability of the cuticle to water vapour probably also ever, the physiological and ecologi- makes it less permeable to other gases like carbon dioxide. cal advantages are less than may be However, if the boundary layer is thick, the resistance to gas indicated by the quantitative difference fl ow added by the cuticle might be of minor signifi cance. On in pool size, because the affi nity for the other hand, in a situation with swift fl ow and thereby a bicarbonate is 2–10 times lower than thin boundary layer, the resistance imposed by the cuticle the affi nity for carbon dioxide [7, 8]. might be of importance. A thin and highly permeable cuticle Despite the obvious advantage of pos- is not necessarily an advantage to amphibious or tempo- sessing the ability to use bicarbonate, no amphibious species able to use bicarbonate have been discovered yet. Amphibious and terrestrial plants species are confi ned to carbon dioxide A use when submerged. The affi nity for carbon dioxide is rather low in these species and it appears that the suc- BC cess of these plant groups in colonising aquatic habitats is dependent on the often very high carbon dioxide Figure 5.3 Most aquatic plants have submerged leaves that are concentrations in streams. When air only 2–6 cell layers thick. Cross sections of leaves of A: Potamogeton and water are in equilibrium with pusillus, B: Potamogeton trichoides and C: Potamogeton pectinatus. respect to carbon dioxide, none of the After [1]. 5 Aquatic plants 73

Box 5.2 Inorganic carbon in water species are able to grow. If the concen- Inorganic carbon exists in water as: 1.0 -- tration of carbon dioxide in the water dissolved carbon dioxide (CO2), bicar- CO3 – –– 0.8 increases 10-fold, which is comparable bonate (HCO3 ), and carbonate (CO3 ), to the average concentration in Danish in a series of equilibria that are control- 0.6 - streams, the growth rate is high and led by pH (see fi gure). The total pool HCO3 comparable to the rate of obligate of inorganic carbon in water is the sum 0.4 submerged species. Upstream and of the three forms. The concentration CO downstream variation in the concentra- of inorganic carbon in water is mainly Relative proportion 0.2 2 tion of carbon dioxide might, therefore, dependent on the geological conditions in the catchment area. High concentra- 0 be a factor regulating the distribution 678910 11 of amphibious plant species. tions are found in areas with calcareous pH soils, and low concentrations are found Oxygen supply in areas with sandy soils. Distribution of inorganic The transport mechanisms of oxygen The pH of Danish streams varies carbon species in water as a in water and the restriction on supply between 6.5 and 8.5 and the predomi- function of pH. rates are similar to those of carbon nant inorganic carbon form is bicarbo- dioxide. However, the demand for nate (see fi gure). oxygen by the plants is lower than the need for carbon dioxide. As a result, the supply of oxygen to leaves and stems of submerged macrophytes The low diffusion rate of gases in The low availability of carbon is generally suffi cient to cover the water presents the submerged plants dioxide, in particular, makes the demands. with another problem not encountered aquatic environment unsuitable for The oxygen supply to roots of by terrestrial plants. The low diffusion plant growth unless the plants employ submerged plants is a quite different rate restricts the rate of gas exchange mechanisms that allow access to story. The characteristics of the soil between the water and the air above. alternative carbon pools or they grow change dramatically in the transient As a result, the oxygen released during at sites with carbon dioxide concen- zone between land and water. On photosynthesis might accumulate trations well above air-equilibrium land a large number of the pore spaces in the water to concentrations up to concentrations. Small lowland streams in the soil are fi lled with air, and the two times the concentration achieved are an example of the latter. Thus, the supply of oxygen to roots of terrestrial when air and water are in equilibrium. active exchange of plant species across plants is good. In contrast to this, the High oxygen concentrations will have the transition zone between stream pores of stream sediment are fi lled with a negative impact on photosynthesis and land, over both short and long water and are usually anoxic – except by partly impeding ribulose-bisphos- time spans, is probably more a result of from the uppermost few millimetres. phate-carboxylase-oxygenase, one of particular physical and chemical charac- As an adaptation to the lack of oxygen the key-enzymes in carbon dioxide teristics of the aquatic habitats than a in the sediment, many submerged fi xation. As a result, rates of net photo- result of specifi c plant traits. When fi rst plants have a well-developed system synthesis might be severely restricted colonised, a more general adaptation of air-lacunae in shoots and roots that in streams developing oxygen super- to the aquatic environment may then allow diffusion of oxygen from the saturation. have developed. However, high carbon shoot to the roots. If the supply of dioxide in the water is most likely a oxygen is insuffi cient the roots will die, Conclusion necessity for the existence of species and this will have a negative impact The most important barrier to the colo- growing in the transition zone between on the anchorage ability of the plants nisation of aquatic habitats by plants is land and water. and thereby on the ability to develop the low availability of carbon dioxide persistent populations. and oxygen in water compared to air. 74 RUNNING WATERS

The beautiful damselﬂ y Calopteryx virgo. 75

6 The terrestrial life of stream-dwelling insects Whereas the biology of the aquatic stages of lotic insects is well studied, detailed knowledge about the activities during their terrestrial life is scarce. These activities include feeding, mating, and ovipositioning. Adults disperse and may thereby colonise not only other parts of the natal stream, but also neighbouring or more distant streams. Together, all these activities are vital for the survival of populations and species. Moreover, adult aquatic insects play an important role in terrestrial food webs. Therefore, there are good reasons for focusing on the riparian zone and other parts of the terrestrial environment when attempting to restore the former quality of anthropogenically destroyed streams.

Life in two different environments Insects constitute a very important The insects probably developed 400 part of macroinvertebrate assemblages million years ago. The early ancestors in Danish and other European streams. were terrestrial, though aquatic groups Especially rich in species are midges, appeared already 290 million years fl ies ( Diptera) and caddisfl ies (Tri- ago. Early mayfl ies (Ephemeroptera) choptera), but also mayfl ies, stonefl ies are known from 286–245 million year- ( Plecoptera) and various beetles are old fossil records, but aquatic groups relatively rich in species. Moreover, as we know them did not appear until several species are widely distributed between 66 and 2 million years ago. and high in abundance, often constitut- Thus, although the insects seem to ing a major part of the macroinverte- have had plenty of time to adapt to and brate fauna. They are therefore well explore the aquatic environment, they studied. We know a great deal about have not totally given up adaptations the immature stages, including life to the former terrestrial life. Generally, cycle, feeding mechanisms, growth, they leave the aquatic environment in distribution, abundance and relations the adult stage of their life cycle. This to other species or individuals of the is true even for elminthid and dytiscid same species. However, our knowledge beetles (Coleoptera) that pupate in about the life of aquatic insects outside moist soil along streams and lakes even the “wet” environment is quite limited. though their larval and adult stages are truly aquatic. These beetles together Communication of adult insects with several aquatic bugs may occa- Stream-dwelling insects leave the sionally disperse over land by fl ying aquatic environment in order to start in order to colonise suitable aquatic the next generation. The sexes need habitats. to locate each other, to mate, and the 76 RUNNING WATERS

Box 6.1 Swarming in adult insects

The males of several adult aquatic insects (e.g. horizontal and vertical movements), here illus- mayfl ies, caddisfl ies, midges) attract females by trated by different caddisfl ies (Leptoceridae and forming swarms. Swarming takes place at particu- Hydropsychidae). One or a few males normally lar sites – markers – in the landscape or above the initiate swarming, whereupon they are joined water surface. Markers may be bushes or trees, by other individuals, sometimes forming dense often located on a hilltop. Each species has its clouds. In Hydropsychidae, males initiate swarm- individual swarming pattern (e.g. lateral zigzag ing by excreting simple chemical substances. movements, vertical circles or ellipses, mixture of Illustration partly adapted from [1].

Hydropsyche pellucidula

Hydropsyche angustipennis

Mystacides azureus

Athripsodes albifrons

Oecetis testacea Ceraclea dissimilis Athripsodes cinereus

Ceraclea annulicornis Ceraclea nigronervosa

Many males of caddisfl ies perform a special fl ight – swarming – to attract females. Patterns and places of swarming are specifi c for each species. 6 The terrestrial life of stream-dwelling insects 77

females to oviposit at an appropriate site. The behaviour of each species during these processes varies considerably, implying that different species have different demands of the terrestrial environment. Several communication methods are used in order to fi nd the opposite sex. Particularly well known is the male swarming of mayfl ies, some families of caddisfl ies and dipterans (e.g., Chirono- Adult stonefl y ( Brachyptera risi) feeding on midae). Swarms of males attract females algae and lichens on a tree trunk. that fl y into the swarms where they are seized by the males, and the couples then leave the swarm to mate (Box 6.1). Figure 6.1 Abdomen of males of different Several caddisfl ies use chemi- adult caddisfl ies (lateral view). Sound sig- cal communication. Females release nals are produced by beating the remarka- species-specifi c pheromones to attract ble protuberances (see arrow and details in males. Such pheromones are found in ventral view) against a suitable substrate. the most species-rich group, Limnephi- A: Agapetus fuscipes. B: Ithytrichia lamel- lidae, and are apparently active over laris. C: Beraea pullata. relatively long distances. Male limnephilids are more active than females in their search for the opposite sex, a selected substrate (stems, branches, body weight about four-fold due to the which may explain why signifi cantly twigs etc.). In some situations the production of eggs [3]. Also the blood- more males than females are usually sounds are produced by wing fl apping sucking behaviour of female blackfl ies caught in light traps. or vibration of the entire body. (Simuliidae) is important for their egg Communication by sound signals The duration of the adult stage production, although a smaller number is used in several groups of terrestrial varies considerably among species, of eggs may be produced in the insects (e.g. caterpillars, cicadas), but families and orders. Mayfl ies are absence of a proper vertebrate blood for decades it has also been recognised especially short-lived and live for a meal. Feeding thus extends the lifespan that male and female stonefl ies com- few hours to a few days. The adults are of adult insects, probably increasing municate by tapping signals. These unable to feed and are therefore com- their potential for dispersal, but also signals are typical for each species, and pletely dependent on energy reserves making them more dependent on suit- it even seems that there are “dialects” accumulated during the immature able shelter against bad weather and among populations. Just recently, the stages. This is also the case for short- predators. The prolonged lifespan of use of vibrational communication has lived members of other orders. Some B. risi and the limnephilid Micropterna been recognised in several families species of caddisfl ies, alderfl ies (Mega- sequax enables these species to over- of caddisfl ies (e.g. Glossosomatidae, loptera), and midges ( Chironomidae) come the regular problem of streams Hydroptilidae, Beraeidae) [2]. There feed on nectar or honeydew excreted drying up during summer, as they are several different types of signals by aphids (Homoptera), which ena- can colonise when these streams are such as drumming, scraping, tapping, bles them to extend their lifespan, fl owing again in autumn. In M. sequax and knocking. The insects produce although the food is not necessary for the lifespan is prolonged by the adults the sounds by moving the abdomen, the production of eggs. In contrast, going into diapause during summer, with its sternal projections located on female Brachyptera risi ( Plecoptera) feed e.g. hiding in crevices beneath the bark segments VI-VIII (Figure 6.1), against on algae, fungi etc. and increase their of trees. 78 RUNNING WATERS

Malaise traps (tents of black, ﬁ nely meshed nylon net) are well suited for studies of lateral dispersal of adult aquatic insects. The insects are led to a capturing container placed in the upper corner of the trap.

Many stream insects do not choose pod number, and 25% and 11% of the of upstream populations. This idea is the oviposition site by random. The biomass [5]. It is also well known that intriguing and has been promoted as females of some blackfl y species are several terrestrial animals including a well-established fact, not least with undoubtedly able to recognise a suit- bats, birds, spiders, and predatory the support of Danish studies of the able spot. Eusimulium vernum only insects eat adult stream insects. Thus, behaviour of mayfl ies and stonefl ies deposits its eggs in forest brooks, the mortality of the mayfl y Dolania [8, 9]. However, there is evidence that whereas Odagmia ornata prefers open americana at a Pennsylvania forest organised upstream fl ight is far from reaches, even if these are just glades stream increased with both increasing universal among stream insects [10]. in a forest [4]. Safely arrived at a densities of a predatory ground beetle First of all, there are several rather proper stream site, the female will and with decreasing densities of alter- contradictory studies indicating that often choose an especially suitable site native prey items [6]. Many aquatic the observed fl ight behaviour is con- to oviposit. Megaloptera and several insects die during their terrestrial life. siderably infl uenced by wind direction, caddisfl ies (especially limnephilids) It is estimated that only about 3% of light/shading, moisture, temperature place their egg masses on branches, adult stream insects survive to return to and riparian vegetation. Apparently twigs and other objects that overhang their native stream. In this way, aquatic documented upstream fl ight may be the water surface. Newly hatched insects may contribute signifi cantly an artefact due to specifi c local condi- larvae will then drop into the stream. to terrestrial food webs, and thereby tions. Calculations furthermore show Other females of certain mayfl ies and to the turnover of organic matter and that infrequent dispersal of randomly caddisfl ies may crawl or swim into nutrients in this environment. fl ying females coupled with improved the stream water to oviposit on dead ability of the immature stages in upper wood, stones, and aquatic plants, while Does an organised upstream fl ight reaches to survive (density dependent) other females are rather unselective take place? may explain why populations are able when they deposit their eggs on the In the mid-1950s, Karl Müller pro- to persist there despite the losses due water surface. posed the existence of a colonisa- to drift [11]. In forest regions close to streams tion cycle among stream insects [7]. Despite this “attack” on an appar- the aquatic insects may represent a Females were supposed to actively ently well-established fact and a large proportion of the fauna. At sites fl y upstream to deposit their eggs in popular theory with strong appeal to located 5 and 150 m from a stream order to compensate for the steady loss the public, stream insects actually do in a northern California forest, adult of their immature stages by drift with disperse upstream. We have observed aquatic insects represented 37% and the fl owing water that would other- this behaviour in several streams on 15%, respectively, of the total arthro- wise lead to the eventual extinction Funen, Denmark’s second largest 6 The terrestrial life of stream-dwelling insects 79

island, where mayfl ies and stonefl ies most original species, but where life 10,000 have been able to extend their distribu- conditions are again suitable for these A tional range by approx. 5 km from one species after a short period of time (e.g. year to the next. However, this might heavy pollution with insecticides). 1,000 just be a result of random dispersal (see Studies using malaise traps (see next section of this chapter). photo on opposite page) in small Danish and English streams have 100 Is dispersal between stream systems shown that most adult stonefl ies and

random? No. of female Trichoptera caddisfl ies stay close to their native 10 The ability of different stream organ- stream [13, 14, 15]. Thus, traps situated 0 10203040 isms to disperse between isolated about 50 m from the stream captured 100,000 streams and whole stream systems 90 % of the plecopterans [13], whereas B deserves attention. This knowledge 95 % of trichopterans appeared in traps may help to explain the way that spe- situated up to 15–20 m from the stream 10,000 cies are distributed nowadays, and [14]. Small caddisfl y species (Agape- to predict how quickly the organisms tus fuscipes, Lype reducta, Silo pallipes) can recolonise restored streams. It dispersed even less, only 1% reaching 1,000 may even have strong implications traps located just 20 m from the stream. for conservation [12]. Differences Dispersal was apparently random 100 No. of female Hydropsychidae among stream insects in their ability to following an exponential decreas- 02,500 5,000 disperse have been observed in several ing function with distance from the Distance from stream (m) other studies. These differences appear stream (Figure 6.2A). Larger species of even between closely related species, caddisfl ies (Plectrocnemia conspersa and Figure 6.2 Lateral dispersal of adult cad- and may depend on, for example, Potamophylax cingulatus) behaved dif- disfl ies from two very different systems. morphological differences. Thus, short ferently and were found more evenly A: Small forest brook on northern Funen, wings, and hence a reduced potential distributed within a distance of 75 m Denmark [14]. B: Detroit River and Lake St. for dispersal, could be one of several from the stream [13, 14]. Clair, Canada [16]. Note the difference in factors explaining the rarity of Swedish Mean lateral dispersal distance of scale for individuals caught in traps, and the mayfl ies and stonefl ies [12]. Trichoptera (Hydropsychidae) was dispersal distance. Knowledge about the dispersal much greater (650–1,845 m) at a large capacity of stream insects may be river and lakeshore in Canada. Thus, extracted from two types of fi eld 1 % of the populations dispersed more studies. First, insects may be caught than 5 km from the water bodies [16], in traps located at various distances showing a much greater potential for relatively fewer and more isolated. from a stream. Depending on the dispersal and ability to colonise distant Consequently, insects inhabiting small distance and trapping device, such water bodies (Figure 6.2B). Light trap- streams only have to disperse a short traps can be attractant (light traps) or ping studies carried out at various distance in order to reach a similar non-attractant ( malaise traps, window Danish localities have also shown that neighbouring stream, whereas insect traps, pan traps). Second, the colo- egg-bearing females of such different- species restricted to larger streams nisation of newly made or restored sized caddisfl y genera as Hydroptila , have to travel much longer distances existing streams can be studied by Polycentropus, and Hydropsyche were to colonise, or to exchange genes with routine long-term monitoring of a fi xed caught up to 8–9 km from the nearest other populations. Despite this, stone- network of sites. The latter group of suitable river habitat. Small brooks and fl ies and caddisfl ies dispersed just less streams include streams that have been streams are naturally abundant and than half of the mean distance between subjected to environmental distur- situated relatively close to each other, small neighbouring forest brooks in bances resulting in the elimination of whereas large streams and rivers are North America [17]. 80 RUNNING WATERS

30 In the Canadian study [16], female signifi cantly further than Heptagenia Leuctra fusca Hydropsychidae seemed to disperse sulphurea (Ephemeroptera). 20 inland just after emergence. Here they In a new man-made Swedish mated with males that often formed stream, blackfl ies were the quickest 10 swarms at distinct landscape mark- colonisers, soon followed by chirono- Heptagenia sulphurea No. of colonised streams ers (bushes, trees). Hereafter, females mids, certain mayfl ies and stonefl ies, 0 dispersed further away from the water whereas beetles, dragonfl ies and 1980 85 90 95 2000 body to fi nd resting sites, where they caddisfl ies arrived much later [18]. might be relatively safe from preda- Successful colonisation of new habitats Figure 6.3 Recent colonisation by adult tors. Then, when the eggs were mature is not only dependent on the distance Heptagenia sulphurea and Leuctra fusca of and ready to be laid, females returned to the nearest inocula. The number of environmentally suitable streams on Funen, to their natal water body to oviposit. possible inoculae and the size of their Denmark, where these species had been However, some females may acciden- populations might be important too. absent for several decades. Note the almost tally – for example under specifi c wind This is well illustrated by the exponen- exponential development in L. fusca, indica- conditions – reach another suitable tial spreading of L. fusca in streams on ting that colonisation progresses concur- water body instead. This could lead to Funen (Figure 6.3). On the other hand, rently with the number of “inoculae”. successful colonisation as some species it seems most unlikely that ten insect produce a large number of eggs. Thus, species that were lost from the Odense it seems that only a few successful Stream (the largest stream on Funen) in females are required to establish new the 1950s would be able to recolonise populations. this stream; the nearest potential inocu- both Danish and Scandinavian streams Such populations are found in our lae in Jutland are few, widely scattered, – occurred in a large river in the Great long-term studies of streams on Funen. and located more than 80–100 km from Belt about 10,300 years ago when the Many of these streams were heavily the Odense Stream. However, one of climate was as warm as nowadays impacted by organic pollution during these species, the caddisfl y Brachycen- [19]. However, one cannot preclude the the period 1960–1980. However, when trus subnubilus, actually re-appeared in possibility that recent distribution of the water quality improved due to autumn 2000 after having been absent some aquatic insects is the result of an extensive treatment of wastewater for 40 years. incomplete dispersal. Anabolia nervosa in the late 1980s, species of various Dispersal, colonisation, and recolo- ( Trichoptera) is very abundant and groups of aquatic insects dispersed nisation should of course be viewed in widely distributed in streams of both over land and successfully colonised the perspective of time scales. There Jutland and Funen, but are not found “restored” habitats situated 4–18 km is evidence that many running water at all on the island of Zealand which is from the “inoculae” (i.e. the source of insects colonised streams from south- separated from Funen by a currently the colonising females) (Table 6.1). The ern habitats soon after the latest degla- 20 km-wide marine Great Belt. In the dispersal ability of the species varies ciation that started about 17,000 years streams on Zealand, A. nervosa is sub- with Leuctra fusca (Plecoptera) and Aga- ago. Thus, more than 20 species of lotic stituted by A. furcata that also occurs petus ochripes (Trichoptera) dispersing caddisfl ies – nowadays recorded in in both Jutland and on Funen, but here

Table 6.1 Overland dispersal of adult aqua- Heptagenia Leuctra Agapetus sulphurea fusca ochripes tic insects in streams on Funen, Denmark. Dispersal distance (km) Median and maximum dispersal distances Median 3 5.5 7 between isolated streams. Differences bet- Maximum 4 10 18 ween H. sulphurea and the two other species No. of observations 7 14 11 are statistically signiﬁ cant (Mann Whitney U-test, P<0,01). 6 The terrestrial life of stream-dwelling insects 81

it exclusively inhabits small lakes and use bushes and trees as swarming sites that the insects fl ew above the vegeta- ponds. The fact that A. nervosa only (see Box 6.1). It therefore makes quite tion thereby escaping the traps. Thus, dispersed a few metres from its home a difference whether the riparian zone the density of vegetation may also infl u- pond during a light trap study on along Danish streams is vegetated by ence long-distance dispersal, thereby Funen supports this hypothesis. either a border of stinging-nettle (Urtica probably explaining the relatively long dioeca), low vegetation of herbs, crops, distances travelled by the insects over The terrestrial habitats of adult more or less scattered bushes and trees, cropped fi elds in the above-mentioned stream-dwelling insects or forest. Cropped areas, in particular, Canadian studies [16]. Studies of the lateral dispersal of adult must be a harsh, inhospitable, and Apparently many factors infl uence stream insects indicate that the major “toxic” site when the crops are sprayed the life of adult stream insects, and the part of the populations stay within with insecticides against pest insects. relationship between these factors can a relatively small zone alongside the The surroundings of a stream no be quite complicated. As knowledge in streams [20]. It is within this zone that doubt infl uence the dispersal of adult this fi eld of research is rather limited, mating, feeding, and maturation of the insects. Species that travel close to the there is a great need for research that is eggs take place. It is therefore impor- ground may experience dense herba- both interesting, challenging, and not tant to obtain knowledge about the ceous vegetation as a barrier. Thus, least, highly relevant for the survival environmental demands of the insects the catch of caddisfl ies at a distance of of stream insects in the fragmented in this so-called riparian zone. 10 m from a Danish stream was greater cultural landscapes around most of Some caddisfl ies choose forests in malaise traps placed in dense beech contemporary Europe. This know- for shelter and rest [6, 21], and many forest without herbaceous vegetation, ledge is also important for evaluating stonefl ies feed on algae and lichens on than in traps located in an open alder the ability of the insects to recolonise the branches and trunks in this habitat swamp or in a fallow fi eld, both having restored streams and the near-stream [3, 13]. The males of some groups (e.g. dense herbaceous vegetation [15]. This surroundings. mayfl ies, alderfl ies, caddisfl ies, midges) result could be interpreted in the way

Danish streams were forested until most of the wood- land was cleared for cultivation. 82 RUNNING WATERS

Capnia bifrons stoneﬂ y nymph. 83

7 Macroinvertebrates and biotic interactions Macroinvertebrates are extremely important in the evaluation of stream water quality. Macroinvertebrates are, however, much more than just bioindicators. They are an essential element for both structure and function of stream ecosystems. Macroinverte- brates are constantly interacting with each other as well as with other stream organisms such as plants and ﬁ sh. These interactions are surprisingly complex and varied, and they have a high overall signiﬁ cance for streams as ecosystems.

What are biotic interactions? investigated. This knowledge is the Stream ecology was previously divided result of signifi cant research efforts into two schools with separate con- within the international scientifi c ceptions of how important biotic community. In Denmark, questions interactions were for the structure and relating to biotic interactions have also function of lotic systems. One school of been addressed within the last decade. scientists viewed streams as being com- This chapter provides examples of this pletely dominated by physical proc- research in which macroinvertebrates esses with very limited possibilities for play a focal role. Before we look at the biotic inter actions between species. The examples, however, it is necessary to other school had the opposite view, defi ne the term “ biotic interactions”. namely that biotic interactions were a Predation and competition are both prominent feature of lotic systems and concepts that relate to biotic interac- that they shaped stream communities tions. Predation includes situations in in a way similar to what is known from which a predator eats a prey (car- other ecosystems such as lakes. How- nivory) and a grazer eats a plant (her- ever, these biological interactions had bivory). Competition for a resource, yet to be discovered and described. often food, can occur between individ- Nowadays the view is more bal- uals of the same species (intraspecifi c) anced: most scientists agree that the or between different species (interspe- truth is somewhere between the two cifi c). With respect to streams, biotic original conceptions. Streams are interactions in the present context also extremely dynamic with large spatial include the effects of the organisms and temporal variability in physical in the riparian zone on the in-stream features and biological elements. As a biota. Examples can be found in the consequence, both types of processes transitional zone between land and are strongly dependent on when and water, and between macrophytes and where stream ecosystems are being macroinvertebrates. 84 RUNNING WATERS

Algal abundance (μg chlorophyll/cm2)

0 0.51.0 1.5 2.0 The land-water transitional zone Alder

To understand the living conditions of stream macroinverte- Eucalyptus brates, it is necessary to focus on the land-water transitional zone. Processes in the terrestrial ecosystem infl uence the Fuchsia amount of food available for the in-stream macroinverte- South Beech brate community. This is most pronounced in small streams Beech that are very closely linked to the surrounding terrestrial Leaf leachate Pine ecosystem. The vegetation on land infl uences the amount of light that reaches the stream surface. Consequently, the Control riparian plant species regulate in-stream primary producers in small streams. In addition, in-stream quality and quantity Figure 7.1 Leaf leachate from the majority of tree species investi- of particulate organic matter (POM) will largely depend on gated had an inhibitory effect on algal biomass, i.e. the biomass was inputs from the riparian zone. It is not possible for detritivo- lower when subject to leaf leachate than on artifi cial leaves without rous macroinvertebrates to affect this donor-controlled con- leachate (controls) [1].

tribution of POM, which constitutes the main energy source in small streams. Another example of interaction across the land-water transitional zone is the leaching of dissolved organic The difference in compounds from tree leaves and their colours was clearly effect on stream algae. When leaves evident in leaf from riparian trees enter the stream, leachates used. various compounds are leached from the leaves into the stream water during the ﬁ rst week of submersion. The composition of the leachate depends on the tree species and has very variable effects on in-stream bacteria and algae. This aspect was investigated in an experiment using diffusion substrates with leachates from six riparian tree species as well as control substrates made from distilled water. Ten diffusion substrates for each species were made by dissolving the leachate in warm agar and subsequently transferring it to small plastic containers. They were subsequently placed in an unshaded Experiment with spring area for two weeks. A piece of artiﬁ cial leaves. nylon netting was mounted on top of Diffusion substrates each container for the algae to colonise with leaf leachates on. This design allowed continuous in a New Zealand leaching of the compounds through spring. the netting to the water. Only leachate 7 Macroinvertebrates and biotic interactions 85

Algal biomass (mg chlorophyll/m2)

0 400 800 1,200 from one tree species stimulated algal tion ran for 44 days in spring when Pre- growth, whereas the others showed growth rates of benthic algae are high- treatment inhibitory effects (Figure 7.1). Beech est and the experimental manipulation and alder, both common in the ripar- should have the greatest impact. At Without insecticide ian zone of Danish streams, were the end of the experiment, the grazer most inhibitory. Compared to other guild was dominated by the freshwater regulatory factors (e.g. light) for stream limpet Ancylus fl uviatilis and non-biting With algae, leaching of mainly inhibitory midges belonging to the subfamily insecticide substances from leaves is probably Orthocladiinae. Especially the occur- of minor importance to the ecosys- rence of Ancylus varied signifi cantly tem function. However, it adds to the between treatments. Only 50 ind./m2 Figure 7.2 Algal biomass was greatest on dependence of small streams on the ter- were found in channels where grazers stones treated with an insecticide solution. restrial ecosystem and will potentially were actively removed using insecti- On stones where macroinvertebrates could limit the energy available for higher cide, whereas nearly 20 times as many graze freely, algal biomass was not different trophic levels such as macroinverte- Ancylus (approx. 900 ind./m2) were than at the beginning of the experiment [2]. brates. Especially during autumn when found in control channels. The number incident light levels increase due to of grazers was successfully reduced leaf fall, leaf leachates might prevent a experimentally, but what about algal 1,000 secondary peak in algal biomass. biomass? There was a signifi cant lower 800 algal biomass (measured as chloro-

Macroinvertebrate grazing on phyll) on stones in control channels ) 600 2 in-stream primary producers with normal grazer densities (Figure Many macroinvertebrates also feed on 7.2). Thus, ambient grazing rates in 400 the streams’ own production of algae spring could reduce algal biomass on (number/m 200 and macro phytes, the most important stones, while algal biomass increased food source being microscopic algae twofold when grazer densities were 0 Grazing macroinvertebrates Grazing macroinvertebrates living on stones and plants. An impor- reduced. This investigation shows that 10 20 30 40 50 60 70 80 tant question in this context is whether grazing macroinvertebrates can regu- Average annual algal biomass macroinvertebrates can regulate algal late the biomass of algae. (mg chlorophyll/m2) biomass through their grazing. The opposite can also occur, namely that algae regulate the density of graz- Figure 7.3 The more algae, the more Interactions between algae and ing macroinvertebrates. Investigations macroinvertebrate grazers. Results for six grazers in six forest streams showed a positive shaded forest streams [3]. In an investigation of Gelbæk, a small relationship between algal biomass and unshaded stream in mid-Jutland, grazer densities, i.e. the more algae, all macroinvertebrate grazers were the higher density of grazers (Figure removed from stones in two experi- 7.3). It is difﬁ cult to determine which forest streams with low algal biomasses mental channels, while they were left of these two interactions are more [4]. Other factors such as disturbance undisturbed in two control channels important: grazer regulation of algal of the streambed during high discharge [2]. Grazers were removed once a week biomass or algae regulation of grazer events are also important for the inter- by submersing stones into a weak density. International investigations action between macroinvertebrates and insecticide solution. At the same time, support the Danish results in that benthic algae. Frequent disturbances stones from control channels were grazers can regulate algal biomass in will reduce algal biomass and macroin- submersed into pure water to mimic open streams with good light condi- vertebrate grazing will consequently the mechanical disturbance of the tions (and consequently algal growth), lose its importance in determining experimental treatment. The investiga- whereas the opposite occurs in shaded, algal biomasses. 86 RUNNING WATERS

Figure 7.4 The a recent investigation in two reaches 25 caddis larvae of Gelså Stream in southern Jutland

(%) Anabolia nervosa indicated the importance of the interac- 20 ) 2 had a signiﬁ cant tion between macrophytes, physical

(g/m 15 impact on the bio- features and macroinvertebrates. One mass of Perfoliate reach had not been exposed to weed 10 Pondweed in spring cutting for over 20 years whereas weed and early summer. cutting took place twice a year in the 5

Grazed plant biomass Plant biomass Later in summer the other reach. Species composition and 0 impact of grazing structural variation in the macrophyte May June July August was insignifi cant community differed between the two 1991 (from [5]). reaches. The reach with frequent weed cutting was completely dominated by Unbranched Bur-reed, which covered approx. 50% of the streambed, while Grazing of macrophytes Plants create habitats for the different plant species were much Grazing by stream macroinvertebrates macroinvertebrates more evenly distributed on the reach on higher plants (macrophytes) was The fi rst thing that strikes you when with no weed cutting for + 20 years. previously considered unimportant. looking at a Danish stream in summer, In this reach, the most frequently However, several Danish investiga- is the lush growth of macrophytes. occurring species was Broad-leaved tions show that macrophytes can Their direct importance as a food Pondweed covering approx. 25% of be an important food resource for resource for macroinvertebrates is prob- the streambed. In addition, in the reach macroinvertebrates. In an investiga- ably limited in most cases, but they are without weed cutting, individual mac- tion in Mattrup Stream, the loss of indirectly one of the most important rophyte stands contained more plant leaves to grazing was estimated for elements for the ecology of Danish species than was the case in the reach Perfoliate Pondweed [5]. Clear effects streams. This is clearly illustrated by with frequent weed cutting. These dif- of grazing were found in spring when an investigation from Suså Stream ferences in the macrophyte community macrophytes have their main growth where 95% of all macroinvertebrates were also refl ected in the macroinverte- period (Figure 7.4). Grazing removed were found within the macrophyte brates: the species richness was higher almost 25% of the plant biomass and vegetation [6]. Macroinvertebrate in samples taken between the plants could delay the development of Per- densities were enormous as more than in the reach without weed cutting foliate Pondweed. When plants have 400,000 individuals per square meter of compared to the reach with frequent built up a large biomass in summer, streambed were found when including weed cutting (Figure 7.5). The same grazing will be of minor importance those living on plants. picture was seen in samples taken and unable to regulate macrophyte In addition to being a suitable beneath the plants. In contrast, no clear biomass. Plants will gradually grow habitat for macroinvertebrates, differences in species richness were out of their problems with increasing macro phytes also interact with the evident from macroinvertebrate sam- biomass, as grazing will be distributed physical environment. By affecting ples taken on the plants even though over a large number of leaves. The current velocities and sedimentation several different macrophyte species most important grazer in Mattrup rates, macrophytes indirectly create a were investigated, an exception being Stream was the caddisfl y Anabolia range of habitats in which the macro- samples taken from the mixed macro- nervosa, which also during summer invertebrates can dwell. Despite the phyte stands (3 macrophyte species) on consumes dead organic matter depos- apparent importance of macrophytes the uncut reach. The mixed stands had ited under the plants. Therefore, the for in-stream macroinvertebrate com- higher macroinvertebrate species rich- macrophytes will indirectly remain munities, only few investigations have ness compared to single macrophyte important for Anabolia. addressed this interaction. However, species. When interpreting the results 7 Macroinvertebrates and biotic interactions 87

Frequent No weed cutting weed cutting 2.4

2.0

1.6 diversity (H')

Macroinvertebrate Macroinvertebrate 1.2

0.8

On/under Bur-reed On/under Water-crowfoot On/under Pondweed On/under mixed vegetation Between plants

Figure 7.5 The greatest macroinvertebrate diversity was found on Brown trout is one of the most important predators on macroinver- the reach where weed cutting had not been undertaken for more than tebrates in temperate streams. 22 years – particularly large differences were found in samples taken between the macrophytes and in the mixed stands [Friberg, original].

from the study in Gelså Stream, the Gammarus is found in the majority of development from March to June in direct impact on weed cutting on the Danish streams and it is probably the the two experimental reaches. While macroinvertebrate communities should most important food source for trout. there was a seven-fold increase in the also be taken into consideration. But That trout can have an impact on number of Gammarus in the two control weed cutting had not been undertaken Gammarus populations has been shown reaches without fi sh, there was only a several months prior to the study in a study of previously fi shless head- two-fold increase in the experimental period and the direct impact can be water streams in the forests Grib Skov reaches with trout (Table 7.1). In addi- considered as minimal. Therefore, the and Rold Skov [7, 8]. In March, juvenile tion to Gammarus, the caseless caddisfl y study demonstrates that macrophytes trout were released into an experimen- Plectrocnemia conspersa (a macroinver- affect macroinvertebrate communities tal reach of 100 m in two headwater tebrate predator) was heavily preyed in streams and it seems to be medi- streams in Grib Skov forest. A control upon by the released trout. ated primarily through changes in the reach, still without trout, was main- physical conditions. tained upstream of the experimental Indirect effects of trout reach in each stream. The released trout on Gammarus Importance of fi sh fed primarily on Gammarus, which In the forest Rold Skov fi ve spring Brown trout (Salmo trutta) is the most constituted approx. 25% of their diet. In brooks without fi sh were investigated. important predator in the majority of addition, the trout were size-selective Due to the shallow depth and various Danish streams. It is also the pri- in their feeding and consumed mainly barriers for migration, no trout were mary game fi sh for anglers and has the largest individuals of Gammarus present in the brooks prior to the been quite extensively investigated. in benthos as well as in drift (Figure experiment. A small enclosure consist- Brown trout affects its prey directly 7.6 and Box 7.1). The preference for ing of a cage with sides of iron netting by consuming them and indirectly by large prey is an advantage for the was placed in each brook for three inducing a behavioural change. Both trout as they obtain most energy using days. Then trout were released into the mechanisms are important in Danish this strategy as long as large Gam- enclosures (four in each) in three of the streams, e.g. in the predator-prey marus are abundant and hence easy to brooks, while two brooks remained as relationship between trout and the catch. Trout predation had a signifi - controls with empty enclosures. The freshwater shrimp Gammarus pulex. cant impact on Gammarus population idea behind this design was to inves- 88 RUNNING WATERS

Box 7.1 Drift

Driftnet in use.

All macroinvertebrates in Danish streams dwell on the time when the search for food commences. the bottom or on plants. However, they are often Catastrophic drift occurs when large numbers of found drifting downstream within the water macroinvertebrates enter drift simultaneously column. The reasons for drift are several and can due to a substantial, external impact, such as a roughly be divided into three categories: behav- chemical pollutant or a sudden surge in discharge. ioural drift, catastrophic drift and background Background drift is the drift that occurs when drift. Behavioural drift occurs as a result of activity there is neither behavioural nor catastrophic drift. relating, for example, to the search for food, or Background drift reﬂ ects accidental dislodgement predator avoidance. This type of drift is primarily of macroinvertebrates. Drift is measured by plac- governed by external stimuli (e.g. light) but also ing a net in the stream trapping macroinverte- by endogenous rhythms controlling, for example, brates that are transported with the current. 7 Macroinvertebrates and biotic interactions 89

Relative increase Stream 1 Stream 2 in number of Drift samples Gammarus

Trout gut content Stream 1 Reach without ﬁ sh 7.3 Bottom samples Reach with ﬁ sh 3.0

Stream 2 Size groups of Gammerus Very small Small Medium Large Reach without ﬁ sh 8.7 Reatch with ﬁ sh 2.3

Figure 7.6 The proportion of large Gammarus was far greater in trout guts than in both Table 7.1 The increase in numbers of Gam- drift and benthos samples. Results were similar for both streams studied [7]. marus was greatest in the two reaches that remained fi shless [7]. tigate indirect effects of trout on the cues in the water released by the trout. macroinvertebrates as the enclosures This interpretation is consistent with Gammarus drift intensity prevented the trout from exerting a fi ndings in international studies on fi sh 0 2,0004,000 6,000 8,000 10,000 direct predation pressure on the prey predation. Interestingly, the large Gam- Day community. Trout were kept in enclo- marus retained their changed behaviour Night sures for three days and then removed, in the third period after removal of the after which sampling was undertaken trout. These fi ndings show that prey for another three-day period before behaviour in streams can be affected by Figure 7.7 Gammarus exhibited a pronoun- the experiment ended. Indirect effects the presence of a predator. Less activity ced night drift pattern already prior to the were measured as changes in mac- during day, as exhibited by large Gam- introduction of trout [8]. roinvertebrate drift by collecting drift marus, reduces the likelihood of being samples approx. 20 m downstream of eaten. However, it also limits the time the enclosures. Already during the spent on food searching to mainly the Proportion of large Gammerus sampling period prior to the intro- night-time period and it therefore has in night drift duction of trout, Gammarus showed energetic costs for Gammarus. 01253 4 With trout pronounced night drift behaviour (Figure 7.7). The signifi cant higher drift Interactions between trout and Before at night compared to day drift refl ects macroinvertebrates During increased activity at night and is most Interactions between trout and macro- After probably an adaptation to avoid trout invertebrates are generally diffi cult to Without trout predation, trout being a visual preda- quantify, the main reason being that Before tor and therefore more effective under the majority of streams has, or have During good light conditions. It is interesting had, a trout population. Therefore, prey After that this behaviour is sustained in communities will already be adapted streams without trout. After introduc- to the presence of a predator and vice ing trout into the enclosures, there was versa. Another diffi culty in assessing Figure 7.8 The proportion of large Gamma- a signifi cant increase in the propor- predator-prey interactions in streams is rus in night drift increased after the intro- tion of large Gammarus in drift during that the estimation of macroinverte brate duction of trout in cages and remained high night (Figure 7.8). They could change quantities (abundance and/or biomass) after the removal of trout. No changes were their behaviour by being less active is probably not suffi ciently precise. found in the two fi shless spring brooks [8]. during the day and more active at Hence, several studies over recent night when trout were present. As the decades have found that fi sh eat more large Gammarus had not been in direct macroinvertebrates than are present in contact with trout, they must have the stream, according to estimations of been capable of detecting chemical prey quantities from the samples taken! 90 RUNNING WATERS

Biotic interactions and the future The present chapter shows that several examples of biotic interactions have been found as part of Danish stream research in the past decade and these examples support fi ndings on the international scientifi c arena. The main problem is how these fi ndings, often obtained on very small spatio-temporal scales, can be linked and put into a common context enabling us to understand processes at the ecosystem level. What, for example, is the importance of macroinvertebrate grazing for the energy fl ow through entire stream ecosystems? – We do not know. Our ability to answer these questions is very limited, as the processes we want to understand are dependent on both spatial and temporal features that are way beyond the scales on which the experiments were conducted. It is a problem because all the major and interesting questions – and answers – are related to the whole ecosystem. We are aiming at being able to assess the impact of various human impacts on the ecological quality of streams, e.g. in the context of the EU Water Framework Directive. Quality is currently almost solely assessed by use of chemical and biological indicators, which provide little or no information on ecosystem functioning. Understanding biotic interactions in streams is interesting from a pure scientifi c point of view, but it could be as important for the future management of streams in Denmark and elsewhere. Studies on biotic interactions in Danish streams are relatively few, and most of the studies involving macroinvertebrates are actually cited in this chapter. There is an apparent lack of studies addressing the importance of competition in structuring macroinvertebrate communities in Danish Mountain streams are systems dominated by physical processes. streams. One reason for the limited number of studies on biotic interactions is probably that the experimental tradition in Danish stream research is relatively weak. In a global context, however, Danish streams are very stable and should There is therefore still a lack of investigations that prove the be ideal for studying biotic interactions. Moreover, there are overall importance of trout, and fi sh in general, in structur- very signifi cant scientifi c challenges in understanding the ing macroinvertebrate communities in streams. This lack interactions between the riparian zone and the stream, and of knowledge comprises both the direct predatory impact between macrophytes and the other biological components. and the indirect effects of behavioural changes. There is no Danish streams are both small and have a rich growth of reason to believe that fi sh are not important in this context, macrophytes making them extremely suitable for this type of but multiple factors will affect biotic interactions in streams. investigations. The scientifi c challenges and the natural set- Biotic interactions are summarised using a hypothetical food ting exist– the question is only whether we seek the opportu- web in Box 7.2. nity and raise our research effort to the level it deserves! 7 Macroinvertebrates and biotic interactions 91

Box 7.2 What controls biotic interactions?

I. Undisturbed 2. Disturbed 3. Polluted

Trout

Macroinvertebrate predators

Grazers/ detritivores

The size of the circles reﬂ ects the relative importance of each element. After [9].

Biotic interactions can be summarised by using a 2. Shows the same food web, but in an unstable hypothetical food web from a small trout stream stream with a very fl uctuating discharge pattern. (1.). Streams have many omnivorous macroinver- The web is inter-connected as before, but the tebrates that feed on different food items. The different trophic levels do not interact as strongly. bottom of the diagram shows the fundamental This is illustrated by an increase in algal biomass food resources consisting of algae (left) and due to reduced grazing pressure. Natural proc- dead organic matter (right). Detritivores and esses linked to physical features of the stream grazers, both primary consumers, use these food therefore defi ne the borders within which the resources. A true interaction between grazers and biotic interactions can take place. A stable stream algae occurs as indicated by the double arrow. is therefore much more likely to sustain complex Grazers feed on the algae and are therefore biotic interactions than an unstable stream. dependent on their productivity, but can, in cer- Human activities can also infl uence biotic inter- tain cases, deplete the algal resource signifi cantly, actions and fundamentally change the function for example when grazers are very abundant and of the ecosystem (3.). In the example, trout has remove more of the algal biomass than can be become extinct due to some kind of pollutant and replenished by the productivity of the algal com- the predation pressure on the large macroinverte- munity. brate predators has been released. As the large Both macroinvertebrate predators and trout macroinvertebrate predators are more effi cient prey upon primary consumers. In this example, predators on grazers and detritivores than trout, trout has a strong regulatory effect (large arrow), both algal biomass and standing stock of dead especially on the large macroinvertebrate preda- organic matter increase as compared to the unper- tors, which they preferred due to their large turbed stream (1. and 3.). size. The caddisfl y Plectronemia belongs to the group of large predatory macroinvertebrates. The trout population is often self-regulatory due to intraspecifi c competition for territories. 92 RUNNING WATERS

Brown trout (Salmo trutta) is a key species in Danish streams. 93

8 Stream fi sh and desirable fi sh stocks Freshwater fi sh communities have always been valuable resources for society, particularly the species that migrate between freshwater and the sea. Historically, eel, salmon and trout were so abundant that good fi shing luck could turn a capable fi sher- man into a wealthy person. Unfortunately, migrating fi sh became threatened by over-fi shing quite early in human history. The more recent construction of dams and other habitat destruction associated with development of towns, industries and intensive agriculture have reduced populations further. The largest Dan- ish river, the River Gudenå, illustrates this historical decline and also the diffi culty of re-establishing healthy fi sh communities once the original populations have deteriorated or, in the case of salmon, become extinct. Efforts are currently needed to be made to effectively restore and manage fi sh communities in streams.

Historical use of stream fi sh Over time almost 40 species of In the milder climate following the last freshwater fi sh established populations Ice Age, new fi sh species immigrated in Denmark. Changes have been seen to Denmark from the ice-free parts of once again during the recent centuries Europe. The immigrants joined the as pollution and physical degradation species already living in the cold lakes have led to the loss of rare and sensi- and streams in SW- Jutland, which had tive species while intentional or fortui- remained free of ice during the glacia- tous introductions have added even tion and had received drainage water more new species. As a result, there are from the glaciers towards the north more fi sh species in Danish freshwaters and east. today than at any other time since the end of the last glaciation.

Frisenvold weir was used for catching salmon in River Gudenå in the late 1800s. 94 RUNNING WATERS

weirs were often constructed in con- It is possible to get some idea of the nection with mill dams. They consisted importance and popularity of differ- of horizontal wooden gratings fi ltering ent fi sh species in ancient history by the falling water, whereupon the fi sh scoring the frequency of place names were led through a drain to the trap associated with fi sh. As number one box, from where the catch could easily on this list, eel is part of more place be retrieved. Whereas there are still names than roach (second place) and about one hundred eel weirs in Danish trout (third place) together. Peak levels streams, the last salmon traps disap- in the abundance of eels must have peared in the early 1900s. The more been reached in the early Middle Age, Historically, eel has been the most important practical fyke nets and pound nets, when lakes, streams and wetlands had freshwater fi sh in Denmark. which are still in use, replaced the fi xed a surface area that was at least three salmon weirs. times greater than today and when Historically, freshwater fi sh most watercourses offered an unhin- remained an important commercial dered passage of glass eel (eel larvae) Through the Middle Ages and up resource for a very long time. In from the sea. until the twentieth century, freshwater regions with many streams such as in fi sh were consumed extensively by mid-Jutland, it was common practice Value of stream fi sh both rich and poor people. Fish were until the 1920s to pay the rent of fi shing Over the last one hundred years the abundant and could be caught in all privileges in natural products – for exploitation of freshwater fi sh has freshwater habitats: lakes, ponds, example a certain annual amount of become more of a sport and leisure streams and ditches. The monks roach. Not all fi sh species were equally activity than a source of energy and brought the most advanced fi shing sought after, however. A suitable size protein. Nowadays, stream fi sh technology available in those days to for a meal was an important criterion. primarily serve as objects of recrea- Denmark from Germany and France. Taste and catchability were other tional fi sheries, but the most attractive In many streams the monasteries important criteria. All these qualities species are still the same. Bream and installed fi shing-weirs that caught the apply to eel, which was very popular. roach have lost their importance in the migrating fi sh with an unprecedented Large, predatory fi sh (piscivores) such freshwater fi shery, as most Danes do effi ciency. When the monasteries as pike, salmon and trout were con- not care to eat these species anymore. were closed after conversion to the sidered to be more delicious but were Instead the recreational fi shery focuses Reformed Lutheran Church, ownership also less abundant than fi sh of lower on the catch of large fi sh, not serving as of the fi shing-weirs was transferred to trophic ranking. The abundance of food, but merely offering a hard fi ght the Royal House. This became the start coarse fi sh such as crucian carp, bream to the anglers before being released of what has since been known as royal and roach made them particularly again. Most meals of eel, salmon and privileged fi shing rights. popular as measures against hunger. trout nowadays come from fi sh reared Different types of fi shing weirs Hunger was a harsh reality in former in aquaculture. were used for catching salmon or eel. days and this had taught the inhabit- Commercial catches of Danish fresh- Salmon traps were wooden construc- ants to exploit the fi sh community in water fi sh are small and are mainly tions placed on “fl at” (i.e. undammed) a relatively sustainable manner. The exported to Germany, France and Italy. currents to catch upstream migrating lesson was simple: in an intensive Only predatory fi sh and salmonids fi sh. They were made of woven willow fi shery there was a larger and lasting (i.e. eel, whitefi sh, perch, pike, trout, twigs placed between solid oak poles, yield by catching the non-predatory salmon, smelt and zander) are caught hammered down into the stream bed. fi sh rather than the predators them- for consumption, while most non- Upstream migrating fi sh were led into selves. Thus, the production of one kg predatory species have disappeared consecutive small compartments, from of predatory fi sh requires in the order from the Danish menu. Even large which they were unable to leave. Eel of 10 kg of prey fi sh. specimens of bream, ide, roach, rudd 8 Stream fi sh and desirable fi sh stocks 95

2,000 2,400 3,500

1,500

(kg) 1,000 Figure 8.1 Annual catch of salmon in the

Catch salmon-weir at Frisenvold on the lower River 500 Gudenå from 1853 to 1913 when the weir was ﬁ nally abandoned. The remaining period 0 shows the total annual catch from 1914 to 1850 1860 1870 1890 1900 1910 1920 1930 1940 1934 when the salmon stock ﬁ nally went extinct. Data compiled from [1] and [2].

and white bream are only caught to a tributed to impoverishing the salmon main spawning grounds for salmon, limited extent by Danish anglers. On habitat, especially for spawning. under several metres of water. A fi sh the other hand, many English, German, From 1898 to 1918 all weekly salmon ladder was built at the dam with the and Dutch anglers travel to Denmark catches at Frisenvold were registered hope that it would allow the passage of to utilise the abundance of coarse fi sh. by A.C. Johansen and J.C. Løfting [1]. migrating salmon and trout. Unfor- The largest of these, the carp, can grow Their study focused on salmon and tunately, the ladder did not work and to a size of 20 kg in Danish waters, and trout and investigated size, age, migra- the salmon stock in the River Gudenå has also become increasingly popular tion and homing behaviour (the ability became extinct, while the trout stock with local specimen anglers. to return to the natal stream). Johansen was able to sustain itself by using and Løfting reported that, economically, suitable spawning grounds in small One hundred years of fi shery studies eel was the most important species, tributaries downstream of the dam. in River Gudenå but that roach was more important In order to solve the problem a new Fishery and fi sh biology have been than eel in Randers Fjord – the estuary fi sh ladder was built and about 500,000 studied for a long time in the River into which the River Gudenå fl ows. To trout fry were annually released into Gudenå, and important information attain an overview of the fi sh commu- the river [2]. However, when E.M. on the development and decline of the nity in the downstream reaches of the Poulsen evaluated the situation in 1935 fi shery is available. For example, in River Gudenå, intensive studies were he recognised that neither the new fi sh 1664 more than 32 salmon traps were conducted with all suitable fi shing ladder nor the restocking programme in operation, whereas this number gear in the lower 50 km of the stream had been successful. Instead he recom- had dropped to 14 in 1833. This channel and in all tributaries between mended the implementation of a four- decline continued and the last salmon the towns of Silkeborg and Randers. point-plan according to which: trap at Frisenvold was commercially The same fi sh species were caught then abandoned in 1907 [1]. The decline of as live there today, apart from salmon, 1) the minimum legal catch size of salmon trapping matches the overall which has become extinct, and zander, salmon and trout should be raised decline of the salmon stock (Figure a recent introduction. from 37 to 45 cm. 8.1). It is known that the original stock 2) restrictions on the use of pound of spring salmon (i.e. multi sea-winter Hydroelectricity and dams nets should be imposed in the lower salmon that enter the river before in River Gudenå River Gudenå, and in the estuary of summer) declined markedly after 1850. In 1920 a 10-metre tall dam was Randers Fjord from April to June, Over- fi shing and sewage pollution established across the River Gudenå 3) the fi sh ladder was to be further were considered the main reasons at Tange in order to produce electric- improved for the decline during the 1800s [1], ity. Lake Tange, the reservoir that was 4) annual restocking of trout fry although straightening and dredging of created above the dam, submerged should be increased to 2 million the river channel most likely also con- agricultural fi elds, farm houses and the individuals. 96 RUNNING WATERS

Following the dramatic extinction lamps in order to direct the migrating because many recreational houses have of salmon in the River Gudenå, the fi sh towards the ladder. Independent been built on the banks. The proposed interest in the infl uence of dams and evaluations in the 1980s and 1990s plan is to improve the river and the turbines on the populations of migrat- revealed, however, that only 5–10% of migration of fi sh by building a new ing fi sh also intensifi ed. In addition the upstream migrating trout actually by-pass channel alongside the lake, and to the hydropower station at Tange, passed through the ladder. In order by letting the majority, if not all, of the fi ve other hydropower dams had to guide more fi sh to the ladder, a 50 water of the River Gudenå through this been constructed in the main channel metre wide metal fence was placed reconstructed river channel. Several dif- of the River Gudenå. To evaluate the across the stream below the outlet ferent models have been suggested. The effects of these turbines on the survival of the turbines. Although the fence by-pass channel will solve the passage of migrating fi sh, C.V. Otterstrøm enhanced the traffi c into the ladder, problem, remove the high mortality conducted experiments with eel and on one occasion it also worked as a risk in Lake Tange and give salmon and trout that were forced to pass through gigantic fi sh trap. While the fence was trout free access to the entire Gudenå the turbines. Several of the turbines lifted for cleaning, trout, salmon and system. Construction costs for some were found to cause severe wounds zander moved in great numbers into of the models have been estimated at in the downstream migrating fi sh, the trailing water below the outlet from amounts as high as 14 million EURO (in particularly the largest individuals [3]. the turbines and when the fence was 2002 prices) or the equivalent of 4 km of The damage depended on the rotation lowered again, the fi sh were trapped motorway. A free river fl ow may turn rate and the number of lamellas and behind the fence. Thus, the obstructed out to be a good investment, improving distance between these in the turbines. fi sh passage at Tange has remained an recreational facilities for local people At one of the hydropower plants, 25% unsolved problem for almost 90 years. and for the tourist industry. of the trout smolt were killed during Currently the future of the salmon passage through the turbines. in the River Gudenå is uncertain; Regulations on the fi sheries The fi sh ladder at Tange was the potential solution is awaiting the A heated debate on the extent of reconstructed several times but only decision of the Danish Parliament. A over-fi shing of trout and salmon was with very limited success [4]. Attempts group of experts have recommended raised in the mid-1980s [4, 5]. Com- were made to use electric current and maintaining the Lake Tange reservoir parisons of size and age data showed that the growth rate of sea- trout had declined markedly when compared with the 1930s. This was attributable to increased fi sh mortality. Investigations revealed that 4 out of every 10 smolt were caught in estuary pound nets [5]. Consequently, in the early 1990s the use of pound nets was banned in Randers Fjord during the main period of smolt migration, i.e. the last three weeks of May. A few years later, in 1991, new investigations showed that the annual migration of trout smolt had more Electrofi shing of than doubled – probably due to an spawning sea trout improved survival of trout at sea and in a typical western thus greater spawning. The spawn- Danish stream ing population had almost doubled (Karup Stream). to 2,700 individuals [6]. A controversy 8 Stream fi sh and desirable fi sh stocks 97

developed between the sea trout experienced a very high mortality rate anglers and the commercial fi shermen during downstream migration towards A Commercial fishermen over the accidental by-catch of sea trout the sea. This was evaluated in 1995 by Anglers 51 % in the pound nets. In 1991, the commer- releasing large numbers of tagged indi- 35 % cial fi shery in Randers Fjord took most viduals of trout and salmon in headwa- of the sea trout catch (approx. 50%), ter streams. Calculations showed that with sport fi shermen also taking a large for every 1,000 individuals released toll (35%; Figure 8.2). When the sport only 18 trout and 220 salmon would Recreational fi shermen realised that they caught a survive the 60-km long journey to fishermen 14 % third of the sea trout and that the unin- Klostermølle where the river fl ows into tended catch of smolt in the pound nets the fi rst large lake, Mossø. On the fur- could be reduced by 90% by lowering ther 120-km long downstream course B Fyke and the nets so that they remained sub- through lakes and river stretches, pound nets merged at low tide, the debate on the only 0.4 salmon smolts are estimated Hook 27 % 35 % fi shing policy silenced [7]. The changes to survive and reach the river mouth did not affect the much more important at Randers. Thus 2,500 smolts would catch of eel in estuary pound nets, so need to be released in the headwaters the ban on pound nets could be lifted. for just one individual to survive and Gill nets be able to leave the river system [8]. 38 % Salmon restocking programme The extensive mortality of smolts in River Gudenå during downstream migration has Figure 8.2 The catch of 2,750 sea trout in In 1998, the municipalities in the town been attributed to the numerous Randers Fjord in 1991 distributed in relation of Randers started a programme to predators in the lakes of the lower to type of fi shermen (A) and fi shing equip- stock one- and two-year old salmon. River Gudenå, and the predator-naïve ment (B) [6]. The aim was to try to restore the behaviour of the stocked smolts. In breeding stock of salmon in the River addition, the lakes reduce the migra- Gudenå. An aquaculture facility was tion velocity and thus prolong the risk built, and eyed eggs were imported of predation by fi sh and birds. Pike are from Ireland, Scotland and Sweden. believed to be the greatest mortality Between 100,000 and 150,000 salmon risk in the lakes while zander are the smolts have been released every year most important predators in the river since 1991. Initially, the salmon smolts channel [9]. were released in all parts of the river Over the past 100 year fi shery stud- A non-migratory brook trout (above) and a as well as at the mouth of the estuary ies have revealed a complex picture migrating sea trout smolt (below). in Randers Fjord. It was hoped that regarding the causes for the decline of improvements of the fi sh ladder at sea trout and the extinction of salmon Tange and the restoration of the breed- in the River Gudenå. It is not possible predation by numerous piscivores in ing grounds in the upstream tributaries to identify a single agent and disregard the lakes and unintentional by-catches would provide suitable conditions for the rest; there are numerous factors, of in estuary pound nets. The poor sur- reproduction. Although more individu- which some may be interconnected. vival of eggs and fry is not only due to als did enter the river and several of The low number of spawning fi sh poor water quality but also to unfa- these succeeded to pass the ladder at cannot be explained solely as a result of vourable physical conditions and lack Tange, there was still no sign of natural over- fi shing, migration obstacles or pol- of breeding grounds. Thus, the recovery reproduction. There were indications lution. The extensive mortality of smolt of prominent salmon and trout stocks that the salmon smolts released in arises from injuries when passing the will require many initiatives at different the upper parts of the river system turbines of the hydropower stations, temporal and spatial scales. 98 RUNNING WATERS

1,000 10 Stocking 800 8

600 6 (tonnes)

(1,000 fish fry) 400 4 Catch 200 2 Stocking Catch 0 0 1920 1930 1940 1950 1960

A ﬁ sh guidance fence and ﬁ sh ladder at Figure 8.3 Improved trout fry stocking schemes since 1937 have generated a consistent the Tange Hydropower Dam on the River relationship between numbers stocked and the annual catch of sea trout in tonnes. Data Gudenå. from the River Gudenå and Randers Fjord estuary [10].

National restocking plans and Electrofi shing was introduced in with local angling clubs. Members electrofi shing surveys Denmark in the early 1950s.An electric participate in the practical work during From a Danish perspective, the River potential is established in the water electrofi shing surveys, gather brood- Gudenå was the fi rst place where it between a stationary cathode and an stock for artifi cial propagation in hatch- was clearly recognised that fi shery anode that is moved by the operator. eries, and release the reared fry. management was urgently needed All fi sh are briefl y paralysed when Electrofi shing also revealed that in order to prevent further deteriora- they come into proximity with the dams and other obstructions in the tion of the commercial fi sheries. It is anode and can then be removed with streams had blocked the access of interesting to note that, at the start at pond net. The fi sh can be identifi ed diadromous fi sh to the headwaters with of the 20th century, the argument for to species, sexed (where possible), the result that many local populations improved management was based on counted, measured and weighed before had either declined or totally disap- the long-term interests of the fi sheries, being returned to the stream. Repeated peared. Therefore, environmentalists not on the endangered stocks them- electrofi shing of the same reach allows and anglers demanded that fi sh passes selves. Although useful as an example, estimation of population parameters be established at all obstacles to fi sh pas- the River Gudenå was not the only such as density, mortality, age struc- sage. A large number of fi shways were place where management schemes ture, biomass and production [10]. built between 1965 and 1985, but their were about to incorporate the idea of By combining visual evaluation of effi ciency was seldom tested afterwards. “sowing before harvesting”. According the stream habitat quality and the trout Because of this uncritical construction, to Otterstrøm’s “Rational Restocking population density, a supplementary most fi sh ladders were not functional. Plan for Trout” implemented in the late stocking programme can be devised, Common errors were to feed the ladder 1930s, the fry should be distributed and in this way an effi cient strategy with too little water relative to the river among all Danish streams according to to optimise the fi sh yield was attained discharge, and inappropriate choices ability of the watercourse to support during the 1970s. The continuous for the location of inlet and outlet of the the growth and production of trout. restocking of trout soon became an ladders. If fi sh follow the main current, Because of this new practice a positive integral part of fi shery management in they easily get misled away from the relationship between stocking and sub- most Danish streams. By 2004, restock- fi sh ladders. Fish ladders are best placed sequent catch of trout emerged around ing of trout has reached a standardised as close as possible to the point where 1940 (Figure 8.3). form that involves close collaboration natural migration is impeded. 8 Stream fi sh and desirable fi sh stocks 99

2,000 50 1,000 Fjord catch 1,600 40 800 Smolt stocking 1,200 30 (tonnes) 600 800 20 (1,000 smolt) Spawning 300 population 400 10

Number of salmon 200 Stocking Catch

Commercial catch Electrofishing Rod catch 0 0 0 1975 1980 1985 1990 1995 2000 1880 1900 1920 1940 1960 1980 2000

Figure 8.4 Annual stocking of smolt and the commercial catch of Figure 8.5 The twentieth century was catastrophic for salmon in sea trout in Denmark correspond. the River Skjern. Commercial catches of salmon in the Ringkøbing Fjord and catches by anglers in the river are shown together with estimates of the total spawning population.

Attractive fi sh populations – omitting landings on the island of the Karup Stream, for example, more During the 1980s the cost of fi sh stock- Bornholm in the Baltic Sea where most than 300,000 hatchery-bred fi sh were ing was included in the Budget of the sea trout derive from Swedish rivers. released between1990 and 2000, but Danish Government. When the National The offi cial statistics show that the more than 90% of the fry were geneti- Rod License – required for angling in commercial catch of trout has doubled cally identical to the wild stock. Among freshwater – was passed in 1993 it was along with the doubling of stocking the sea trout virtually all were origi- decided that the income should be used intensity (Figure 8.4). This suggests nal wild fi sh, while the proportion of for stocking fi sh and for administra- that the commercial fi shery has ben- hatchery-bred fi sh was larger among tion and research on fi sh management efi ted from the stocking programme the resident brown trout, which spend to improve the stocks. So far, resources as well as from overall improvement their entire life in the river. It appeared have mainly been used for rehabilitation of the environmental quality of the that while sexually mature dwarfs of of sea trout and, to a smaller extent, on freshwaters. hatchery-bred fi sh were capable of eel and salmon, refl ecting the principle New DNA-analyses have shown fertilising some of the eggs of spawn- species of interest for the recreational that there is a drawback to the massive ing sea trout, they themselves were not fi sheries in freshwater. stocking programme. Wild populations able to migrate successfully to the sea. of resident and migrating trout face the Apparently hatchery fi sh are geneti- Trout risk of competition and hybridisation cally defi cient or poorly adapted to the The widespread release of small trout with hatchery trout bred in aquaculture complex and fi ne-tuned life cycle of sea over the past 30–40 years has increased facilities. DNA analyses have revealed trout which involves migration to the the number and size of Danish trout that in some rivers the stock consists sea, growth in the sea for one to several populations, although it is diffi cult to entirely of “pond trout” or their genes years and return to the home river to separate this effect from that of habitat are mixed with the original wild stock. reproduce. Therefore, if the objective is improvements and improved sewage In several other rivers distributed to ensure a native self-reproducing fi sh treatment. Only a few watercourses are throughout Denmark, however, the stock it is a bad idea to release foreign currently completely devoid of natu- wild stocks are small, but still geneti- fi sh and fi sh from hatcheries. It is also rally reproducing trout. The impact of cally intact and they are obviously bad business as the introduced fi sh trout stocking can be seen in the land- much more effi cient at reproducing generally have much higher mortality ing of sea trout in the Danish harbours than the introduced “pond trout”. In rate and are very ineffi cient at repro- 100 RUNNING WATERS

48 River Storå during the 1950s 44 River Storå during the 1930s Figure 8.6 “Genetic distance” between 34 River Storå 1913 River Skjern 1989 salmon stocks in different Danish, Swedish 49 (River Ätran) and Scottish (River Conon) 66 49 River Skjern during the 1950s River Skjern during the 1930s rivers [11]. 76 River Skjern 1913 92 River Sneum 1913 River Varde 1913 95 River Gudenå 1913 Genetic analysis was made on River Gudenå during the 1930s old salmin scales. River Conon River Ätran

ducing compared to the well-adapted the river (Figure 8.5). The other large of the original salmon stocks have wild stock. Moreover, in certain rivers in Jutland offer the potential to recently been found. circumstances released foreign fi sh can support salmon stocks in the future. In When local stocks have become become strong competitors for the wild the Varde Stream, Sneum Stream, Ribe extinct, it is better to release fry from stock in the stream and severely reduce Stream and Vidå Stream a comprehen- “neighbouring” stocks rather than use the survival of the fry and young of sive survey for surviving individuals of fi sh from more distant populations the wild stock. As a result, much fewer the native salmon stock has been con- [12]. Fish from the native populations native fi sh develop and migrate to ducted before stocking is continued. are adapted over many thousand years the sea, thereby diminishing both the Captive breeding and release of fi sh of to the characteristic environmental commercial catch and the size of the the native stock is preferred to that of (e.g. temperature regime and water reproductive population returning to foreign stocks. chemistry) and biological (e.g. prey, the river for spawning. DNA analyses of preserved salmon competitors and predators) conditions scales have shown that salmon caught at the local sites and they consequently Native salmon stocks in the River Skjern are derived from perform better than foreign fi sh. Since 1993 salmon have been released the native stock. The genetic composi- in streams that have been known tion of the living population closely The decline of eel historically to have supported natural resembles that found in analyses of Eel populations have experienced salmon populations. The objective of DNA from scales of fi sh caught in 1913 a dramatic decline during the 20th the salmon rehabilitation plan was and in the 1930s and 1950s [11]. Similar century (Figure 8.7). The collapse of to re-establish self-sustaining stocks DNA analyses of salmon caught in the Danish eel fi shery resembles that in the streams. However, only in the other Danish rivers in the past, have observed in the rest of Europe; catches River Skjern has a natural, reproducing also shown that the genetic distance have diminished and the numbers population of salmon survived with between stocks from different rivers of glass eel reaching the European certainty. There are indications that refl ects the geographical distance coast have declined year after year. the long-term decline of salmon in the between the rivers (Figure 8.6). It Nonetheless, the intensive fi shery has River Skjern has now come to an end, has therefore been recommended to continued in Denmark. A study in the as a combined result of effi cient fi shery search carefully for remnants of the Sound between Denmark and Sweden regulations in the estuary Ringkøbing native populations in the Danish rivers revealed that the fi shery is so effi cient Fjord, greater water exchange through and, if they still exist, to use these in that of every fi ve silver eel captured, the sluice separating the North Sea and the breeding programme whenever tagged and released, one is recaptured the estuary and improved passage of possible. In two streams, namely Ribe within a few weeks. The eel fi sheries fi sh across the physical obstructions in Stream and Varde Stream, remnants industry has continued to show an 8 Stream fi sh and desirable fi sh stocks 101

Figure 8.7 The populations e.g. by causing inter-sex 1,600 number of eel is and disturbing sexual maturation. Cur- Yellow-bodied eel declining steeply, rently there is a lack of knowledge of 1,200 here shown in terms the long-term and wide-scale impor- of the total commer- tance of these potential threats. 800 (tonnes) cial catch of yellow The immigration and introduction Silver eel and silver eel in of foreign and native species of animals

Catch 400 Denmark. and plants can infl uence the composition of fi sh communities. One example 0 is the establishment of predaceous 1975 1980 1985 1990 1995 2000 mink populations in many Danish streams. Large numbers of this Ameri- can mink have escaped from fur farms, and could be having a strong impact alarming lack of concern and responsi- tionary principles in the management on stream fi sh communities. Also the bility for these issues: “As many silver of European eel, as they have not as quantitative infl uence on fi sh commu- eel as possible should be captured yet been able to identify, with certainty, nities of the decline and recent recov- before they migrate out of the national the key factors controlling the size of ery of fi sh-eating cormorants, grebes, coastal waters”. The attitude is based the eel stock. The overall recommen- egrets, seals and otters is unknown. on the belief that a suffi cient number dations are to maximise the growth Beaver has recently been re-introduced of silver eel will succeed to spawn and of eel populations and minimise all to one Danish stream system and their thus secure an unchanged recruitment controllable mortality factors. To attain dam-building behaviour may have of glass eels, independent of the fi shing a sensible management plan, however, a strong impact on stream habitats, intensity. The data do not support this an agreement is needed between all obstructing the free fl ow of water and belief. On the contrary, the number of European eel- fi shing countries. This forming wetlands and pools. A decline glass eel that reached Western Europe agreement has not been reached and as in the abundance of fi sh species prefer- in 2000–2002 was only a few percent of yet no master plan has been imple- ring fast fl ow in the stream channel the 1950–1980 numbers. Sustainability mented. There are several reasons for (e.g. trout, salmon and minnow) may has never been on the agenda of the eel this lack of determined action. The be expected whereas species preferring fi sheries and the result has been severe most important reason is insuffi cient sluggish fl ow in pools and backwaters over-fi shing. knowledge of the key factors regulat- (e.g. bream, roach, perch and pike) are In Denmark, silver eel has no legal ing the population size of eel. likely to increase. minimum size or a closed season. Ongoing urban expansion poses Fishing for silver eel is even permitted New threats another threat to streams and their fi sh in river outlets, the transitory zones New threats to fi sh populations in communities. Abstraction of ground- where all other fi sheries are normally streams will emerge in the future but water reduces discharge in streams prohibited. Therefore, if the decline are diffi cult to foresee. The growing use and increases the risk of small streams of European eel fi sheries continues, of genetically modifi ed organisms for drying up temporarily. Conversely, a marked change in the attitude is food may well transfer to the aquacul- effi cient drainage of streets, buildings, needed. The seriousness of the situa- ture industry. It is possible that geneti- car parks, etc. quickly returns precipitation has grown from being a question cally modifi ed farmed fi sh could escape tion over large areas to the streams. As of preserving the size and health of to the streams and infl uence wild popu- a result, temporal discharge patterns the eel population, to preserving the lations through hybridisation. Also become more erratic with associated future existence of the European eel as chemical pollutants, their degradation extreme variations in current veloc- a species. Fishery biologists have rec- products, and hormone-like substances ity, erosion/sedimentation, tempera- ommended the enforcement of precau- constitute a risk to the fertility of wild ture and oxygen concentration in the 102 RUNNING WATERS

A sea trout smolt result in a more natural abundance and tagged with a radio diversity of the semi-aquatic fl ora and transmitter placed in fauna. More terrestrial insects along the the abdominal cavity streams would serve as food for most with the antenna fi sh species. Another goal is to use the sticking out through potential of wetlands to reduce the dis- a small hole. solved and particulate nutrient loading of streams. Eutrophication, anoxia, and fi sh kills are regrettably still common streams, generating greater physical New options in many Danish streams, lakes and and physiological stress on animal Several of the Danish plans initiated coastal waters. and plant life. For example, increased after the UN conference in Rio de Fish research has been introduced to sediment input could reduce salmonid Janeiro in 1992 also provided potential many new technical options. The pos- egg survival in spawning grounds. In for improvements in stream habitat sibility of examining the undisturbed low-gradient Danish streams, sedi- and biodiversity. Among these were behaviour of fi sh in their natural ment deposition represents a common initiatives to increase forest area, shift environment has improved consider- danger to the survival of fi sh eggs from conifers to a mixture of broad- ably. By equipping individual fi sh buried in the stream bed. leaved trees and change to a more with minute radio transmitters it has Fish stocking can also represent a extensive management of the forest to become possible to follow every move- threat to wild populations. Stocked fi sh allow variations in landscape types and ment, muscle contraction and heartbeat are often raised in aquaculture facili- to encourage biodiversity. Currently, while the fi sh lives relatively undis- ties and as such are poorly adapted to there are many physically homogenous turbed under fi eld conditions. Small wild conditions. When released, they streams in conifer plantations, these act instruments can record the continuous can potentially inter-breed with wild essentially as drainage ditches and have changes in light, temperature, oxygen populations, diluting the native genetic low diversity because of their uniform and current velocity in the environ- strains. Furthermore, fi sh mortality can morphology, acidic waters and tend- ment. New automatic instruments can be increased in wild populations due to ancy to dry up during summer. One count the numbers, measure the size greater fi shing intensity on the artifi - of the proposed plans is to bring water and take pictures of fi sh each time they cially enhanced populations. In Den- back to the forests and to aim towards a pass obstructions during upstream or mark, the stocking of trout smolts at more natural hydrology, closing drain- downstream migrations in the streams. the mouth of streams and in the marine age tubes and ditches and allowing The “fi sh counters” follow the princi- coastal waters has increased due to the swamps, lakes, springs and streams to ple that the fi sh intercept infrared light pressure from sports fi shermen wanting re-establish or improve. Forest streams beams as they pass, and thereby they to catch more sea trout along the coast. have a major advantage because most generate a silhouette which is used to The stocked smolts have never experi- are headwaters with little or no inten- determine the direction of the move- enced the taste from “a home stream” sive management e.g. weed cutting and ment, the swimming velocity, the size unlike natural trout or smolt released in dredging. Therefore, forest streams may and the identity of the species. Every the streams. The spawning migration of be ideal for attaining stable spawning picture can be saved electronically and such straying sea trout is thus probably grounds for salmonids. a programme for pattern identifi cation less precise. The genetic consequence There is also potential for the can discern plant debris and stream of straying on the native populations is improvement of fi sh communities in plants from the true fi sh counts. Such not well known, but the risk of out- streams in the open land if intensive automatic fi sh-counters should greatly breeding is probably highest in small agricultural practices along streams improve the control of fi sh passages streams with vunerable populations. are abandoned and natural wetlands and offer quantitative measures of the are given room to develop. This may numbers of migrating fi sh. 8 Stream fi sh and desirable fi sh stocks 103

Trends in recreational fi shery A male salmon elec- In 1995 the United Nations’ Food and trofi shed from the Agriculture Organisation, provided a River Skjern to be code of conduct for responsible fi sher- used as broodstock ies, declaring that: “States and users of for producing 1 and living aquatic resources should conserve 2-year old salmon aquatic ecosystems. The right to fi sh smolts for restocking. carries with it the obligation to do so in a responsible manner so as to ensure effective conservation and management ies for families with small children in satisfy many confl icting interests at the of the living aquatic resources”. friendly settings. In Denmark put- one time. All severe impacts on streams So far, Germany is the only Euro- and-take fi sheries do not take place and their fi sh communities have been pean nation requiring sport fi shermen in streams, but only in ponds or lakes accepted, including straightening, to follow a short course and pass an where the fi sh are unable to escape. dredging, input of sewage and polluted exam before they can receive a license Catch-and-release is another scheme drainage water, water abstraction for to fi sh in natural waters. The applicant that is expected to spread in the future. domestic use and for fi sh ponds, and must show some basic knowledge on The idea is to release fi sh that are hydro-electric power production. Dams fi shery biology, environmental condi- caught and to do it as gently as pos- and other obstructions have isolated tions, handling of fi sh and fi shery sible. The objective is to optimise the many fi sh populations within small, legislation. The objective of the course- chance of an individual fi sh surviving restricted areas. Where migrating popu- exam-license system is to improve the several catches and to allow it to grow lations have gone extinct, obstructions management of aquatic ecosystems. to an attractive size. Therefore, catch have increased the vulnerability of non- In Germany the green movement has and release is adopted in closed waters migrating populations to pollution. focussed on the well-being of animals and with non-migratory fi sh so that the Even though Danish streams as a and this has generated a push for same person or members of the same whole have more fi sh species today improving the ecological knowledge fi shing club have a greater chance of than in the past, local species richness and the proper behaviour of sport fi sh- catching trophy fi sh. Anglers fi shing is probably much lower because of the ermen [13]. In France 500 schools have for carp mostly use catch-and-release, overall deterioration in the physical and been established that offer courses in but the method has also been used chemical condition of stream habitats. recreational fi shery [13]. Both examples in stream fi sheries on brown trout. In order to restore the species richness suggest that courses and insight into There are strong indications that this of stream communities it is essential to sport fi shery will probably grow in will be the only acceptable manage- ensure the re-establishment of habitat Europe in the future and that education ment strategy if fi shing on threatened heterogeneity, unhindered passage and may stimulate interest in the sport. populations such as the salmon in the relatively stable stream beds. A common practice in recreational River Skjern is to be permitted. The Fortunately, it is still possible to fi sheries is to release hatchery-bred North Atlantic Salmon Conservation improve the relationships between salmonids as a basis for the fi shery in Organisation has published a pam- environmental objectives and the small, often excavated, artifi cial lakes. phlet describing how to treat a hooked practical management of streams and Such put-and-take lakes have spread salmon to ensure its survival upon their fi sh populations. Richer and more throughout Europe during the 1980s release. robust fi sh populations can develop by and are still growing in numbers. Most means of well-founded, co-ordinated, Danish put-and-take lakes are stocked Fish management and targeted efforts. with rainbow trout weighing from 0.5 The management of fi sh populations to 3 kg. Apparently there is a strong in Danish streams over the centuries demand for safe recreational fi sher- illustrates an approach that tried to 104 RUNNING WATERS

Dense vegetation of pondweed in lower River Gudenå. 105

9 Water plants past and present How has the condition of the environment developed during the last 100 years? This is an important question because the conditions of the past inﬂ uence those of the present, and knowledge about this development makes it possible to evaluate the causes of changes in species richness, species composition and environmental quality. Here we compare old and contemporary studies of the Danish stream ﬂ ora.

Early studies of water plants However, the botanists 100 years ago As early as the late 19th century there had a particular interest in pondweed was a strong interest in natural history species [1, 3], and therefore we will in many European countries. This focus on the historical development interest resulted in several studies of within this group. the stream fl ora in Denmark [1, 2]. We have searched early publications and The stream environment 100 years ago excursion reports from 1870 to 1920 for Maltreatment is probably the best word information on the Danish stream fl ora, to describe the use and management and additionally included information of streams during the last 100 years. from the main Danish herbarium at The biological and ecological value of the Botanical Museum of Copenhagen. the streams was completely ignored Based on this material we have com- in the 19th and 20th century in favour of pared old species lists from 27 streams using the streams for drainage, land distributed across the country with reclamation and diversion of wastewa- species lists obtained from our own ter. However, improved wastewater study on the vegetation in 208 streams treatment and an increased focus on in Denmark carried out about 100 years the ecological quality of the stream later. In this comparison of old and environment, during the past 10 to 25 recent studies, 13 reaches are identical years, has been a basis for improving and therefore directly comparable. the fl ora and fauna in streams. We consider the comparisons What kind of maltreatment have reliable because the species identifi ca- Danish streams, along with other tion has been carefully undertaken and lowland streams around the world, suf- based on almost identical identifi cation fered during the past century? The four keys in the past and present studies. most important impacts on streams 106 RUNNING WATERS

This reduces plant species richness because only a few disturbance-tolerant species with high recolonisation ability can tolerate such conditions [5]. We would expect a decrease in species specialised to live at permanently high water velocities or at consistently low velocities e.g. in backwaters. Several of the latter species are also common in lakes and were previously widely distributed in streams, but have now nearly disappeared. Already in the late 19th century weed cutting was practised in some streams, but in most streams the practice was not introduced until after 1920. Weed cutting and dredging of Today’s streams the stream bed by use of machinery receive more intensifi ed between 1960 and 1990. organic pollution Nowadays the vegetation is cut at and nutrients from least once a year in nearly all Danish towns, farms and streams, and the situation is similar in cultivated areas many other lowland areas in the world. compared to 100 In this highly disturbed environment years ago. we may expect the dominant species to be those that are able to quickly regrow and form new plant stands from and their vegetation are: straighten- rest have lost much of their physical individuals or fragments that have ing of stream meanders, weed cutting, heterogeneity (see Chapter 1). The survived the cutting. It may also be organic pollution and eutrophication. channelised streams quickly divert species that are effi cient dispersers and More than 90% of the Danish water from the catchment resulting in that quickly establish new populations streams have been channelised in order increased differences between winter on a bare stream bed that thrive under to drain wetlands, and the proportion and summer fl ow, and increased fl ow such management. Effi cient dispersal is similarly high in other countries after heavy rainfall. and settlement capability, and high with high agricultural land usage. The loss of area and habitat het- growth capacity are therefore necessary Most channel regulation was carried erogeneity in the streams would be characteristics of plant species surviv- out during the period 1850–1965. The expected to have lead to a reduction in ing a disturbed stream environment. most severe regulation, using of heavy the total number of species both in the Species with high growth capacity machinery, occurred towards the end entire stream and in some individual have an advantage in a nutrient-rich of this period. Regulation meant that reaches. This prediction is based on stream environment. Lowland streams streams lost their meanders as well the universal ecological concept that: are often nutrient-rich because of the as the natural sequence of riffl es and species number and diversity increase intimate contact with the surrounding pools. At the same time many of the with increased area and habitat diver- land, from which nutrients are released smallest creeks and streams were sity [4]. High-fl ow variation results continuously. The nutrient concentra- piped. Several thousand kilometres in alternating erosion and sediment tions are especially high in streams sur- of streams have disappeared and the deposition zones on the stream bed. rounded by pastures or arable fi elds. 9 Water plants past and present 107

Water is only retained in the stream for a short time due to the continuous current, and only a small proportion of the nutrients are used in the stream. Most nutrients are transported to downstream lakes and coastal areas. Moreover, the streams receive much more organic pollution from urban and agricultural areas today than 100 years ago. At that time, the human population was lower, there was less sewage, and the use of fertiliser was minimal. It is a well-known fact that most lowland freshwater systems in Europe have been subjected to extensive organic pollution over the last 100 years. In Denmark, the impact on streams was probably most severe between 1950 and 1970. Wastewater pollution has certainly impoverished the fauna in the streams, but the plants were also affected because species able Callitriche species to survive on unstable mud sediment (starwort) are facing periodic anoxic conditions were common in Danish favoured while more susceptible spe- streams today. cies declined in abundance. However, the situation has improved with enhanced wastewater treatment start- Prediction and testing of changes vegetation have been reduced, and 2) ing in the late 1970s. The strategy that we adopted for it has become more common for a few Eutrophication of Danish streams analysing vegetation changes in Danish species to dominate plant communi- has been widespread and extensive streams during the past century was ties. We also predict that: 3) eutrophic both in terms of dissolved inorganic based on known changes in physical species, with high growth rates and ions and total nitrogen and phospho- and chemical characteristics of streams high dispersal capacity have become rus. Nowadays, nitrate concentrations during the period, and established more frequent compared to species in streams are typically 1–4 mg N/l, ecological rules. This allowed us to with slow growth and poor dispersal probably 5–10 times higher than 100 set up testable predictions of expected capacity, due to high disturbance and years ago [6]. Typical concentrations of changes in the vegetation that could be eutrophication in streams. Species total phosphorus are also high (0.1–0.3 rejected if they were not supported by expected to have increased are those mg P/l [6]). The nutrient-rich condi- the observed situation in present-day with an ecological ruderal strategy tions favour nutrient-demanding, streams. (R-strategists sensu Grime [7]). Species eutrophic species, and restrict oligo- Stream conditions have changed expected to have decreased in number trophic species, either because the latter towards smaller surface areas, reduced during the period include species with cannot tolerate high nutrient concen- habitat variation, high disturbance low competitive ability that are only trations or they are out-competed by and unfavourable stream bed charac- able to dominate under oligotrophic eutrophic species. teristics. Therefore, we predict that: 1) and stable conditions (S-strategists), species richness and diversity of the large species with high competitive 108 RUNNING WATERS

Box 9.1 Ecological plant indices

Eutrophication indices A morphology index assesses the architecture of pondweed species and thereby their ability to grow up through Fennel Pondweed P. pectinatus the water column and ramify at the surface to cope with Broad-leaved Pondweed turbid waters. Species are divided into three groups based P. natans on their height: 1 is species <0.5 m, 2 is species 0.5–1.0 m, Curled Pondweed 3 is species >1.0 m. The ability to ramify near the surface P. crispus or develop fl oating leaves is scored as: 1 for species with- Perfoliate Pondweed P. perfoliatus out ramifi cation or fl oating leaves, 2 is intermediate, 3 for Shining Pondweed species that often ramify or easily develop fl oating leaves. P. lucens Grass-wrack Pondweed A trophic index describes the species distribution in P. zosterifolius Long-stalked Pondweed streams according to their trophic level. The 16 pondweed P. praelongus species in Danish streams are ranked from 1, which is the Red Pondweed species in the most oligotrophic streams, to 16, which is P. alpinus the species in the most eutrophic stream. It is not known Flat-stalked Pondweed P. friesii conclusively whether this ranking accurately refl ects the Lesser Pondweed nutrient demands of the species. P. pusillus Blunt-leaved Pondweed Dispersal index P. obtusifolius The ability to disperse at a regional scale, within and Sharp-leaved Pondweed P. acutifolius between stream systems, by means of stem fragments, Slender-leaved Pondweed specialised dispersal organs and seeds, is assessed on a P. filiformis three-level scale:1 describing low, 2 intermediate and 3 Opposite-leaved Pondweed high dispersal ability. The ability to disperse on a more P. densus local scale, within a site, is assessed by the ability of the Various-leaved Pondweed P. gramineus root system to disperse underground: 1 is species without Bog Pondweed a large root system; 2 is species with root systems <0.5 P. polygonifolius m in diameter; 3 is species with root systems >0.5 m in diameter. A collective dispersal index is the sum of the 1896 0246810 regional and the local dispersal indices. 1996 Number of sites

Figure 9.1 Frequency of 16 pondweed species at 13 sites surveyed in 1896 and 1996 [9].

ability and those demanding relatively position has changed towards a higher By means of the literature we have high nutrient concentrations and low frequency of disturbance tolerant and found three species traits among the disturbance (C-strategist). eutrophic species with high dispersal pondweeds that may be described by It is possible to determine the spe- ability. To do that, we need to express numerical indices (Box 9.1). A trophic cies richness and species composition the ability of the different species to index developed for British streams in the stream vegetation today and 100 tolerate disturbance and eutrophica- ranks the species according to their years ago and to test if any changes tion. Then we can compare the ecologi- occurrence in streams with low and have taken place. It is more demanding, cal strategies of species in previous and high nutrient concentrations [8]. An however, to examine if the species com- contemporary communities. index characterises the morphology 9 Water plants past and present 109

25 A

20 Relative Freqency 15 1896 1996

Sharp-leaved Pondweed 0.11 0 10 Red Pondweed 0.41 0 * 5 Curled Pondweed 0.41 0.30 Number of species 1996

Opposite-leaved Pondweed 0.07 0 0 Slender-leaved Pondweed 0.15 0 0 5 10 15 20 25 Flat-stalked Pondweed 0.33 0.02 * Table 9.1 Relative 25 Various-leaved Pondweed 0.15 0 frequency of 16 B pondweed species Shining Pondweed 0.52 0.03 * 20 in 1896 and 1996 Broad-leaved Pondweed 0.74 0.18 * at 27 and 208 sites, 15 Blunt-leaved Pondweed 0.23 0 * respectively. For Fennel Pondweed 0.48 0.18 * species marked with 10 Perfoliate Pondweed 0.59 0.15 * * there is a low 5

Bog Pondweed 0.04 0 probability (p<0.05) Number of species 1996 Long-stalked Pondweed 0.41 0.08 * that the decrease 0 0 5 10 15 20 25 Lesser Pondweed 0.33 0.01 * in occurrence from 1896 to 1996 is Number of species 1896 Grass-wrack Pondweed 0.56 0.01 * coincidental. Figure 9.2 Number of pondweed species (A) and total number of submerged species per site (B) at 13 and 11 stream sites, respectively. The sites were surveyed in 1896 and 1996. The 1:1 line indicating no change in species occurence is shown.

of the plant species and their ability to grow up through the 1896, but only 7 species in 1996 (Figure water column and ramify near the water surface in order to 9.1). The number of pondweed species capture the maximum amount of light. Plant species more and the total number of submerged capable of withstanding the poor light conditions usually species per site have also decreased accompanying eutrophication, score higher. A dispersal markedly over the period (Figure 9.2). index evaluates the dispersal ability of the species locally If we include all the locations that were in their habitat and regionally over greater distances within surveyed, 16 pondweed species were the stream system. Because experimental work is lacking we found across 27 sites in 1896, but only were unable to assess the growth rate of the species and the 9 species at 208 sites surveyed in 1996 direct tolerance for weed cutting. (Table 9.1). In summary, 7 out of the original 16 Drastic decrease in pondweed vegetation species were not even observed in 1996, We found substantial differences in species richness and 8 had decreased in abundance, and composition between the past and present pondweed vege- only one species (curled pondweed) tation. Pondweed species were very common in streams 100 has maintained the same abundance. years ago but most are rare today. At the 13 survey locations that are identical in both studies, there were 16 species in 110 RUNNING WATERS

Table 9.2 Indices Dispersal Morphological Trophic Ranking Ranking index index index 1896 1996 describing dispersal Common species in 1996 ** ability, morphological ability to cope Bog Pondweed 4 4 6 10 12 with turbid water, Fennel Pondweed 6 6 16 12.5 14.5 and trophic ranking Perfoliate Pondweed 4 4 14 15 13 of 16 pondweeds. Curled Pondweed 6 4 15 10 16 The ranking of Broad-leaved Pondweed 6 6 5 16 14.5 species in 1896 and 1996 ranges from Rare species in 1996 1 (most rare) to 16 Bog Pondweed 4 4 1 1.5 4 (most common). Flat-stalked Pondweed 2 2 12 7.5 10 In 1996, common species were found Blunt-leaved Pondweed 2 2 2 6 4 at 50% of the sites Grass-wrack Pondweed 2 3 10 14 8.5 while rare species Shining Pondweed 4 4 13 12.5 11 were found at less Various-leaved Pondweed 4 4 3 4.5 4 than 5% of the sites Lesser Pondweed 2 2 7 7.5 8.5 investigated. Red Pondweed 4 4 4 10 4 * indicates signi- Sharp-leaved Pondweed 2 2 9 3 4 ﬁ cant differences Slender-leaved Pondweed 5 2 11 4.5 4 between the groups. Opposite-leaved Pondweed 6 2 8 1.5 4

Greater tolerance for disturbance Today, the abundance of the pondweed Species with high tolerance for and eutrophication species is in better accordance with disturbance and nutrient-rich condi- We tested our third hypothesis, on their dispersal ability than 100 years tions have become more common trajectory changes in the stream vegeta- ago (Table 9.3). while nutrient-poor species and spe- tion during the past 100 years, using The pondweed species most cies susceptible to disturbance have two methods. We compared the mean abundant in the streams today have a decreased in frequency or disappeared disturbance and eutrophication index higher morphology index than the spe- completely. The group containing the values between frequently-occurring cies that have disappeared or become nutrient demanding species with high species in the past and contemporary rare (Table 9.2). Remember that this dispersal ability (R-strategists) includes surveys, and between rare species in morphology index describes the ability the four pondweed species: Potamoge- the two surveys. Moreover, we studied of the species to grow tall and ramify ton pectinatus, P. perfoliatus, P. crispus the relationships between indices and near the water surface as a response to and P. natans. The nutrient-poor species the frequency of the species in the past bad light conditions. Species relative (S-strategists) which have strongly and contemporary surveys. abundance is more strongly correlated decreased includes Potamogeton prae- The pondweed species occurring with their trophic index today, than longus, P. alpinus and P. ﬁ liformis. The most frequently in the streams today in the past, while the correlation with group of slow growing species that do have a higher dispersal index than the their morphology has not changed not tolerate disturbance includes Pota- species, which have disappeared or signiﬁ cantly (Table 9.3). mogeton polygonifolius, P. zosterifolius decreased in abundance (Table 9.2). and P. lucens. These large broad-leaved 9 Water plants past and present 111

Current predominant stream plants Ranking Ranking 1896 1996 At present Danish stream vegetation is dominated by species Dispersal index 0.15 (0.57) 0.37 (0.15) that prefer nutrient-rich conditions and tolerate disturbance. Morphological index 0.58 (0.02) 0.49 (0.06) Such species include Potamogeton crispus and P. pectinatus, which are ranked as number 9 and 11, respectively, on Trophic index 0.43 (0.10) 0.62 (0.02) the list of the most frequently occurring plants in Danish Table 9.3 The frequency of pondweed species shows stronger streams (Table 9.4). Sparganium emersum (Unbranched Bur- relationships (the fi gures are correlation coeffi cients) between reed) and Elodea canadensis (Canadian Waterweed) are the dispersal ability and nutrient demand in 1996 than in 1896. Small most common species and the predominant vegetation in values in brackets indicate a low probability that the relationships about 50% of the Danish streams. Callitriche species (Star- are coincidental. worts, mainly C. cophocarpa and C. platycarpa) and Ranuncu- lus peltatus (Pond Water-crowfoot) are also placed higher on the list of abundance than the pondweed species. The most frequent stream plants all have high growth capacity and good dispersal ability. Sparganium emersum tolerates cutting well because it grows from the basal meristem, which usually survives cutting, or from the vigorous underground parts buried in the stream bed. Elodea canadensis disperses naturally by shedding stem fragments and this strategy is enhanced when the population is disturbed by cutting. The heavy fragments fl oat with the current but easily sinks to the bottom to form roots and regrow at new sites downstream. Water Mint Detached stem fragments of species of Callitriche and Ranunculus also disperse effi ciently with the current until caught by stones or other obstacles. They then develop adventitious roots from the stem and become established species were previously very common, but are now quite on the stream bed Moreover, Callitriche and Ranunculus can rare. For example, we only found P. zosterifolius at one of 208 disperse by seeds. The importance of seed dispersal is only stream localities examined in 1996 while it was present at known from a few studies, but it is likely to be signifi cant for 56% of the localities examined in 1896. dispersal over long distances and for establishment in new There are other observations that show the substantial territories, while dispersal by stem fragments is more impor- infl uence of disturbance on the aquatic plant communities. tant to sustain the populations within the colonised stream. Past surveys found that, not only were there many pondweed species in the streams, but that they also formed large Many rare stream plants continuous beds. Baagøe and Kølpin Ravn reported that the Among the 20 pondweed species growing in Danish streams broad-leaved pondweed species in 1895 created macrophyte only 4 species are still common. About 10 species have expe- stands up to 60 m long in the two largest rivers in Denmark rienced a substantial decrease over the past 100 years. Many [2], but these were smaller where weed cutting had taken of these have also become rare in lakes due to eutrophication place. In contemporary streams with frequent weed cutting [11]. Among the rare pondweed species in Denmark, we can we never fi nd such large stands of pondweeds. New studies list 5 species as very rare, namely: Potamogeton zosterifolius, confi rm that stream vegetation is richer in species and forms P. rutilus, P. acutifolius, P. densus, P. coloratus. Unfortunately, a more complex canopy, where it has not been cut for over P. polygonifolius, P. gramineus, P. alpinus and P. fi liformis now two decades compared to reaches that are cut once a year face a high risk of also being included among the very rare [10]. species. 112 RUNNING WATERS

Macrophyte species Frequency Species belonging to plant genera (%) other than the pondweeds have also Unbranched Bur-reed become more rare in streams. This is 75.1 Sparganium emersum Rehman particularly true of species associated Water-Starwort with still or slow-fl owing water such as 69.7 Callitriche sp. Myriophyllum spicatum (Spiked Water- Canadian waterweed 60.6 milfoil), M. verticillatum (Whorled Elodea canadensis L. C. Rich Water-milfoil), Ranunculus circinatus Lesser Water-parsnip 57.7 (Fan-leaved Water-crowfoot), Cerato- Berula erecta Hudson (Coville) phyllum demersum (Rigid Hornwort) Water Forget-me-not 51.9 and the submerged form of Stratiotes Mysotis palustris L. aloides (Water-soldier). These species Branched Bur-reed 51.9 Sparganium erectum L. were previously common in the lower parts of streams and in backwaters, and Water-crowfoot 36.1 Batrachium peltatum (Schrank) Presl. were favoured by their ability to dis- Blue Water Speedwell perse from lakes to streams. Now these 32.2 Veronica anagallis-aquatica backwaters have disappeared and the Curled Pondweed lower parts of the streams often have 30.2 Potamogeton crispus L. turbid water fi lled with phytoplankton Floating Sweet-grass from upstream lakes. Eutrophic lakes 29.3 Glyceria fl uitans (L.) R. Br. have lost their submerged vegetation Fennel Pondweed 18.3 and thereby their ability to enrich the Potamogeton pectinatus L. lower stream parts with plant prop- Broad-leaved Pondweed 17.8 agules such as stem fragments, short Potamogeton natans L. shoots and seeds. Pond Water-crowfoot 17.8 A number of other species are Batrachium baudotti (Godron) F. Schultz threatened by extinction. This applies Flowering-rush 16.8 Butomus umbellatus L. to Luronium natans (Floating Water- plantain), Alisma gramineum (Ribbon- Perfoliate Pondweed 14.9 Potamogeton perfoliatus L. leaved Water-plantain) and Oenanthe Common Water-crowfoot fl uviatilis (River Water-dropwort). 13.5 Batrachium aquatile (L.) Wimmer Luronium natans and Oenanthe fl uviatilis Water Mint receive great attention because they, 12.5 Mentha aquatile (L.) Wimmer along with Potamogeton acutifolius, only Watercress 11.5 occur in Western Europe. These species Nasturtium sp. have experienced a drastic decline that Water Horsetail 11.5 threatens their global survival. Equisetum fl uviatile L. Fan-leaved Water-goosefoot 11.1 Possibilities of re-establishment Batrachium circinatum (Sibth.) Spach What hope is there of recreating a more Table 9.4 Relative frequency of submerged plants at 208 stream diverse aquatic plant community, once sites in 1996. water quality has been improved and streams reaches remeandered, when so many species are now so rare? In fact no one has been able to answer this question, but we can make some 9 Water plants past and present 113

100

50 (%) 10

Local frequency 1

–0.5 –0.5 1 5 10 50 100

Stream vegetation Regional distribution (%) is dominated by species that are Figure 9.3 Relationship between the regio- tolerant to distur- nal distribution of 57 stream species in Den- bance and nutrient- mark in 1996, and the mean local frequency rich conditions. at the sites where the different species occur.

general refl ections based on known risk of disappearance from the already ing lost habitats, even if the environ- theories and relationships. occupied habitats”. mental conditions were to improve. First we have found a general When the number of favourable While physical and chemical condi- positive relationship between the habitats for water plants becomes tions can be re-established by technical geographical distribution of stream limited in streams as well as in lakes, means, the re-establishment of biologi- plants across Denmark and their the local frequency of the species may cal diversity is much more diffi cult. local abundance (Figure 9.3). The also decrease in habitats not directly The maltreatment of streams, lakes relationship means that species with infl uenced, because immigration from and wetlands has put many species in a restricted geographical distribution suitable neighbouring localities is a marginal position that unfortunately also generally have a low abundance restricted. Thus, when the frequency may turn out to be permanent. at the sites where they do occur. On the of the species decreases in habitats In this chapter we have shown other hand, species with a widespread because of environmental deteriora- that the composition of stream plant distribution among the streams usually tion, the distribution of species may communities is not simply related to have a high abundance at sites where decrease in the entire country, and in current environmental conditions but is they are present. The relationship can the long term, over the entire distri- also shaped by the legacy of historical be explained by differences in their bution area. Essentially core areas of development. Furthermore, the conse- ecology. Species able to utilise com- a species distribution, i.e. the areas quences of past abuses are seldom fully monly-occurring resources, or a broad where the species attain the highest reversible. So yes: The past century range of resources, have widespread abundance, can ensure the supply of has made a huge difference both to distributions, while species dependent individuals to remote satellite areas the present and future vegetation in on rare resources or a small range of so that the species may also maintain streams. resources have restricted distributions. a population there. The possibility of An alternative explanation is proposed supplying the satellite areas with prop- by the relationship between frequency agules will decline when the popula- and dispersal ability, which proposes tions in the core areas decrease. that: “High local frequency provides At present, many Danish water opportunities for colonisation of plants grow in so few habitats and in many habitats, attaining a widespread such small populations that statistically distribution and thereby reducing the they have very little chance of reclaim- 114 RUNNING WATERS

The adult mayﬂ y Ephemera danica. 115

10 Improvements ahead for macroinvertebrates ? During the 20th century, anthropogenic pressures, such as wastewater discharges, reduced the biodiversity of Danish streams. However, over the last 30 years, large sums of money have been invested in wastewater treatment and stream restoration in an attempt to improve the ecological quality of these streams. But what is the present state of the freshwater biology? Have we achieved improvements in environmental quality as a result of the investment? In this chapter, attention will focus on the macroinvertebrate fauna of Danish streams, and it is evaluated whether the ecological quality of the streams has improved in recent years. Finally, it is discussed how further changes are expected to affect the ecological condition of the streams.

Macroinvertebrate fauna in Danish 1,600 samples of mayfl ies, stonefl ies reference streams – how was it and caddisfl ies were collected. These before? insects were preserved in alcohol and Currently, the majority of Danish thereby kept for future study. In some streams are continuously exposed to watercourses, such as Funder Stream, one or more anthropogenic pressures Århus Stream and the lowermost parts (such as wastewater discharges, regula- of River Gudenå and River Skjern, tion, weed cutting, drainage, water there was particularly intensive sam- abstractions, heavy metals and the use pling by the zoologists Esben Pedersen, of pesticides). The crucial question is J. Kr. Findal, Hj. Ussing and Carlo therefore, has the macroinvertebrate F. Jensen. In these streams we have communities changed as a result of particularly detailed information about these impacts? the macroinvertebrate fauna before There is historical information on wastewater discharges, regulations the macroinvertebrate communities in the stream channel and changes in of Danish streams dating from the catchment use occurred. beginning of the 20th century. During Therefore, it can be seen that we the period 1900–1950, collections were have some knowledge about the obtained from many streams. In 1916 pre- disturbance state (or reference and 1917, the zoologist, Wesenberg- condition) of the macroinvertebrate Lund, initiated an investigation of a fauna in Danish streams. A number number of streams situated in different of species are only known from the parts of the country [1]. The macroin- museum collections and have never vertebrate fauna was sampled by local been found since that time (Table 10.1), professionals and amateurs; about even though recent collections have 116 RUNNING WATERS

Previous detections 1900–1960 Last Present status 1990–1999 detection Mayﬂ ies Baetis digitatus [3] Widespread ( Zealand, Jutland) 1955 Extinct? Baetis muticus [3] Widespread (Zealand, Jutland) 1918 Extinct? Heptagenia longicauda [4] Hadsten, Lilleå Stream 1912 Extinct? Siphlonurus lacustris [4]Funder Stream 1918 Extinct? Rhithrogena germanica [5, 6]Grejs Stream, Jeksen Brook, - Højen Brook River Gudenå, Bangsbo Stream, Varde Stream, Hansted Stream Baetis buceratus [3] River Skjern 1956 River Storå (one induvidal in 2004) Stoneﬂ ies Perlodes dispar [4] River Gudenå 1955 Extinct?

Dinocras cephalotes [4] Grejs Stream, 1949 Extinct? Table 10.1 Selected Orten Brook near Varde mayfl ies, stonefl ies Brachyptera braueri [4] Grejs Stream - River Gudenå and caddisfl ies (one induvidal in 1991) registered in the Siphonoperla burmeisteri [4, 7] Widespread in Jutland - Vidå Stream (River Skjern, Varde Stream, Ribe Stream, Red Data Book of River Gudenå, Kolding Stream) threatened and rare Caddisfl ies plants and animals Micrasema setiferum [8] Funder Stream, River Gudenå 1917 Extinct? in Denmark [2]. Hydroptila forcipata [9] Lellinge Stream 1907 Extinct? Some species are Hydroptila occulta [9] Lindenborg Stream, Gjern Stream 1948 Extinct? now believed to be extinct. Other spe- Wormaldia subnigra [10] Sæby Stream - Mattrup Stream, Øle Stream cies are still present Lasiocephala basalis [5, 10] Århus Stream - Mattrup Stream, Åsbæk Brook but were previously Odontocerum albicorne [10] Occasionally - Rare more widespread. been much more comprehensive, both years ago. Here we still fi nd some Most of the species that are now countrywide and at the previous sam- species of caddisfl ies, which are now rare or have already disappeared from pling sites. Some rare species, now only rare in Denmark, but were previously Denmark (Table 10.1) are associated found in a few streams, are present in found in the Funder Stream, the Århus with medium or large streams. These the museum collections from a greater Stream and other similar watercourses species represent a variety of differ- number of streams [3, 4, 10]. It is (Table 10.1). But generally, these spe- ent functional feeding groups and expected that these species were much cies are no longer commonly found in life strategies. However, for many of more widely distributed and abundant Denmark. these species, Denmark represents the than today. It is important to protect these northern or north-western limit of their A few Danish medium-sized and exceptional streams that have an distribution in Europe. In contrast, some larger streams have only been undisturbed macroinvertebrate fauna some species, which are believed to minimally impacted over the years. because such streams are scarce not be extinct in Denmark, can still be These streams give us another oppor- only in Denmark, but also in most found in isolated populations north of tunity to obtain knowledge about the other parts of the North European Denmark. natural composition of macroinverte- lowland. The continued existence of The elimination of rare species from brate communities. these undisturbed macroinvertebrate stream systems or even larger areas Mattrup Stream, a tributary of communities is also important as they represents a loss of biological diversity the River Gudenå, is an example of a act as a potential source of colonists for at the local and the regional scale. Such medium-sized stream with a macro- nearby streams as and when they are losses are in many cases diffi cult or invertebrate community as it was 100 restored. impossible to reverse because many 10 Improvements ahead for macroinvertebrates in Danish streams? 117

40 (%) The most important pressures on 30 Danish streams Why do most Danish streams not 20 comply with the quality objectives? The regional authorities responsible 10

Distribution of sites for performing the biological assess- 0 ment of streams have concluded that 1234567 nearly 12,000 km streams do not have Fauna classes acceptable macroinvertebrate communities. Furthermore, the regional authori- Osmylus fulvicephalus. The larvae live under Figure 10.1 Moderate ecological quality ties have evaluated the reasons for the moss, bark and stones in the transitional (fauna class 4) characterised Danish streams unacceptable ecological quality in the area between water and air. in 2002 (1,051 stream sites investigated) [15]. streams (Figure 10.2). The main impact is believed to be wastewater from sewerage systems and scattered dwellings macroinvertebrate species are unable to out Denmark [14]. Since 1999 the and it is estimated that the ecological disperse over long distances. number of stream sites has increased condition of 6,000 km of Danish streams Despite the changes in macroin- to 1,053 and the same method has been is unacceptable as a result. The discharge vertebrate composition over the last used annually at these stream sites. of wastewater to streams is refl ected century, the dominant species found These surveys have found that the in the concentration of organic matter in unimpacted streams a century ago intermediate quality class, fauna class 4, (BOD5) in the stream water. Wastewa- are probably the same as are found indicating a moderately impacted site, ter from about 99% of the wastewater in streams today. However, some of is the most common in Danish streams treatment plants is biologically treated the most sensitive species of mayfl ies, (43%). Streams having a diverse fauna, [16] and the BOD5 content in the outlet stonefl ies and caddisfl ies are no longer with a number of sensitive species of sewage treatment plants is typically found in Denmark at all or they have (fauna classes 5, 6 and 7), account for 2–3 mg/l. However, diluted wastewa- survived in only a few streams. 37%, while heavily-impacted streams ter discharges during rainfall events (fauna classes 1, 2 and 3) constitute are believed to result in many streams Today’s streams; how is the ecologi- about 20% (Figure 10.1). remaining in an unacceptable condi- cal quality? More than 20,000 km of Danish tion. The amount of oxygen-consuming For the last 20–25 years, the Danish streams are subjected to quality organic matter from wastewater has regional authorities have assessed objectives, set by the regional authori- been calculated since the Nationwide environmental stream quality using ties, based on the macroinvertebrate Monitoring Programme was initiated biological stream assessment based on community. In 2000, about 45% of the in 1989 [17]. The discharge of organic macroinvertebrates. Since the mid- Danish streams complied with their matter from wastewater treatment 1980s, 10–15,000 assessments have been objectives [15]. Compliance typically plants and industrial plants was reduced made every year. The same assessment requires an unimpacted or only slightly by 92% and 80%, respectively, during methods have not been consistently impacted macroinvertebrate commu- the period 1989–1999. The discharge of used (Box 10.1) across all regions or nity (fauna classes 5, 6 and 7). Unfortu- organic matter measured as BOD5 from over the 25-year period. As a conse- nately, some regional authorities accept about 350 fi sh farms was reduced by quence, it has been diffi cult to make fauna class 4 in slow-fl owing streams more than 50% during the same period countrywide and temporal compari- with poor habitat quality. Overall, the although, locally, biological conditions sons of ecological quality. ecological quality of Danish streams is may still be unacceptable [17]. The Danish Stream Fauna Index far from the reference conditions asso- Currently, the most important (DSFI) was therefore introduced in 1998 ciated with fauna classes 6 and 7. wastewater problem is the discharge at 444 stream sites distributed through- of mechanically treated wastewater 118 RUNNING WATERS

Box 10.1 Biological stream assessment in Denmark

Biological stream assessment made by the Ministry of Agriculture were the offi cial method Four different assessment methods based on macroinverte- for biological stream assessment from 1970 to 1998. brates have been used in Denmark between 1970 and 2000. The basic principle is that many species of larvae of stone- Viborg Index fl ies, mayfl ies and caddisfl ies can only thrive under certain The Viborg Index also uses seven pollution degrees [12]. environmental conditions, e.g. fast fl owing oxygen-saturated Sample collection is standardised. The sample sorting and water, dense riparian vegetation cover. In contrast, some spe- identifi cation is performed in the laboratory. The pollution cies of midge larvae and worms tolerate most pressures and degree is obtained by an objective procedure. For example, can survive in polluted or physically disturbed streams. Based a species list can only give one single pollution degree and on the complete list of macroinvertebrates from a stream site is therefore not subject to different interpretation by two or an index value can be calculated (pollution degree or fauna more persons. The Viborg Index is a biotic index where the class). This value expresses the ecological quality (see below). combination of specifi c indicators are used together with the number of selected diversity groups to calculate the actual index value. Some county authorities have based their interpretation of the guidelines of the Ministry of Agriculture on the Viborg Index from the mid-1980s until 1992.

Danish Fauna Index (DFI) Danish Fauna Index is a further development of the Viborg index [13]. The sensitivity of DFI has been adjusted in order to clarify the distinction between moderately and strongly Nymphs of the stonefl y Perlodes microcephalus require clean polluted streams. The index values are named fauna classes water and a stony substrate. instead of pollution degrees because the composition of the macroinvertebrate community is not only affected by the organic pollution, but also by the habitat quality, ochre Guidelines from the Ministry of Agriculture 1970 compounds, hazardous substances and the water fl ow. These guidelines were originally inspired by the Saprobic Danish Fauna Index has been used as the offi cial biological system [11] used mainly in central Europe since the beginning assessment method at 222 stream sites under the Nationwide of the 20th century. The guidelines briefl y describe the distinc- Monitoring Programme from 1993 to 1997. The DFI has also tion between seven different pollution degrees (I, I-II, II, II-III, been used by county authorities who previously used the III, III-IV and IV). Today the Saprobic system is mainly used in Viborg index. Germany and Austria. The method requires laboratory sorting and identifi cation to species level. The species list is used Danish Stream Fauna Index to calculate the Saprobic value by use of a special formula. In Danish Stream Fauna Index (DSFI) is a further development Denmark there has been a tradition for compiling a fauna list of the DFI [14]. A minor change has improved the separation in the fi eld, and then based on the fi ndings, making a subjec- between fauna classes 4 and 3. The fauna classes in DSFI are tive judgement on which of the seven pollution degrees is denoted by numbers from 1 to 7; an index value of 1 cor- appropriate for the site. Therefore, the guidelines from the responds with a strongly impacted stream, whereas an index Danish Ministry of Agriculture differ considerably from the value of 7 corresponds to unimpacted or slightly-impacted more traditionally used Saprobic system. The weakness of the streams. In 1998, the Danish Stream Fauna Index was intro- method recommended by the Ministry of Agriculture is that duced as the new offi cial method for biological stream assess- it is not reproducible. Different people can interpret a species ment. In 1998 the method was used at 444 stream sites in the list in different ways. And the same person can change his Danish Aquatic Monitoring and Assessment Programme. In interpretation as more experience is gained. The guidelines 1999, the number of stream sites was increased to 1,053. 10 Improvements ahead for macroinvertebrates in Danish streams? 119

from scattered dwellings into small Figure 10.2 Rea- streams [15]. This is believed to be the sons for non-comp- Lake outlet main reason for unacceptable ecologi- liance with quality Reduced flow 2% Area serviced by 9% sewerage systems cal quality in about 3,000 km of Danish objectives in 12,000 Toxic substances 15% streams (Figure 10.2). The wastewater km Danish streams 2% Freshwater fish farms 4% discharged into small streams is [16]. The three Ochre inadequately diluted and therefore has colours in the fi gure 11% Illegal discharges 4% a negative impact on the macroinver- illustrate the pres- tebrate communities. Consequently, it sures from organic is necessary to improve the treatment pollution (light Regulation and Single houses of wastewater from sparsely built-up green), habitat maintenance 23% 27% areas. Initiatives have now been intro- degradation (dark Low gradient streams duced, and most regional authorities green) and other 3% have selected a number of small stream pressures (blue). catchments where the implementation of improved treatment of wastewater from scattered dwellings will reduce fi ne-material. This is usually due to still maintains the macrophyte com- the discharge of organic matter and reductions in water velocity associated munities in a modifi ed condition. A thereby aim to improve the ecological with the bank-to-bank dimensions of further reduction in weed cutting or, quality of the streams. At the moment, the streams, which have been increased alternatively, a complete termination this work is limited to these selected as part of the stream maintenance. is a both desirable and feasible option areas, and it is expected to take many Weed cutting reduces the amount that would improve the habitat quality years before the majority of stream of vegetation periodically but it also and result in a higher number of more catchments receive similar treatment changes the species composition of natural biological communities. programmes. the macrophyte community [18]. As a Other pressures leading to unac- An additional problem in Danish result, the macroinvertebrate commu- ceptably poor ecological stream quality streams is a general reduction in the nities are reduced in abundance and include draining of pyrite-rich soils habitat quality of some streams as the species composition is changed. leading to ochre pollution, and water compared to natural, unimpacted Revision of the Watercourse Act in abstraction leading to changed fl ow streams. Habitat quality can be reduced 1982 resulted in the introduction of regimes and a general reduction in as a result of stream regulation, heavy- more environmentally-aware stream water velocities. handed stream maintenance and the management methods. Maintenance erosion of nearby agricultural areas was changed from a near-total removal Have the macroinvertebrates com- leading to increased inputs of nutrients of macrophytes and eventually a munities improved in recent years? and sediment into the stream channel. removal of the bottom sediments, to a The answer can be found by compar- As a result of these impacts on habitat more gentle maintenance with mac- ing the general ecological quality in quality, 3,000 km of streams do not rophytes being cut only in the main the streams 10 years ago with that comply with the quality objectives channel with a signifi cant proportion of today. We know that the targeted [15]. As mentioned above, the major- of plant material being left untouched. effort of the counties and the munici- ity of Danish streams are more or less Consequently, the fi ne sediments are palities has resulted in a signifi cant affected because of previous regula- typically no longer removed from the reduction in polluting substances in tions and present stream maintenance stream bottom [18]. This has resulted streams. It is mainly the reduction in methods. These pressures maintain the in an improvement in the macroinver- organic pollution that is responsible streams in an unstable condition. In tebrate communities and also trout for the improved ecological quality. In many cases stone and gravel substrates abundance [19, 20]. However weed the 1960s and 1970s new wastewater have been removed or covered with cutting is now more frequent and treatment plants were built or existing 120 RUNNING WATERS

(mg/l) 5

2 Figure 10.3 Mean 1 BOD5 content in 0 streams has decre- Adult female of the Organic substances in streams 1975 19801985 1990 1995 2000 2005 ased from 1976 to dragonﬂ y Calop- 2004. teryx.

plants were improved. Consequently, has now stabilised at about 2 mg/l. A Viborg Index of II-III (moderate there has been a signifi cant reduc- This level is still higher than that found ecological quality) is the most common tion in the most severe pollution from in unimpacted streams where the BOD5 for both periods. The most signifi cant urban areas. Rationalisation and the content is typically about or below 1 difference is that the proportion of shutdown of small dairies and slaugh- mg/l, but a BOD5 level of 2 mg/l is heavily-impacted stream sites (III, terhouses in small villages have also not a limiting factor for the majority of III-IV and IV) has decreased from 28% eliminated these sources of pollu- macroinvertebrate species. to 18% with a corresponding increase tion for many streams. Furthermore, The interesting question is whether in the proportion of un-impacted, pollution from direct discharges of there are more pollution-sensitive mac- weakly-impacted and moderately- seepage water from manure etc. into roinvertebrates in Danish streams now impacted stream sites (Table 10.2). This the streams has now been stopped due than 10–15 years ago? This question general improvement in the ecological to an active effort by the county and can be answered using information quality is found in both small (< 2 m municipal authorities. from 651 stream sites situated in the wide) and larger streams (> 2 m wide). Unfortunately, there are no reliable County of Southern Jutland. It would In the small streams there has been a national estimates of the extent of these have been preferable to use results signifi cant reduction in the number of reductions in pollution inputs. But it from various sites in all of Denmark, heavily-impacted streams and a cor- was during this period that the fi rst but the use of different assessment responding increase in the number of major efforts were made to improve the methods has, as previously mentioned, moderately-impacted streams. In the environment. The only reliable values made it diffi cult to make both inter- larger streams, where generally, the for the pollution loads from cities, regional and temporal comparisons. ecological quality is better, there has industry, fi sh farms etc. are from after Fauna lists from the County of mainly been a reduction in moderately- 1989 when the Nationwide Monitoring Southern Jutland are here expressed impacted streams corresponding with Programme under the Action Plan on as pollution degrees using the Viborg an increase in the number of weakly- the Aquatic Environment was initiated. Index method [12] which resembles impacted streams. In the mid-1970s, two Danish coun- the method currently used, the Danish Several Danish counties have regis- ties started analysing the content of Stream Fauna Index [14]. Since 1986, tered this general improvement in the

BOD5 in a number of streams (Figure the county has used the Viborg Index ecological quality although different 10.3). The results show that the water as an aid when estimating the level assessment methods have been used. quality has gradually improved up of pollution in accordance with the The tendency is similar: the majority of until the late 1980s and provides a offi cial guidelines provided by the streams previously classifi ed as heavily clear documentation of the effect of the Ministry of Agriculture (Box 10.1). The impacted can now be classifi ed as only strong environmental measures intro- Viborg Index has been calculated using moderately impacted. In the medium- duced during the 1970s and 1980s. In the macroinvertebrate fauna lists for sized and large streams, the improve- many streams, the BOD5 concentration the periods 1986–1987 and 1995–1997. ments are also similar to those found 10 Improvements ahead for macroinvertebrates in Danish streams? 121

Viborg Index 1986 1995–1997 streams where migration has appar- Previous initiatives have almost ently occurred from one watershed to eliminated illegal contributions of farm I (undisturbed) 12 18 another. wastes. Wastewater from cities is now I-II 21 26 II 121 146 In the River Storå, the largest stone- more effectively treated, with a result- II-III 313 347 fl y in Denmark, Perlodes microcepha- ing improvement in water quality. III 79 69 lus, was for many years only found However, the ecological quality is still III-IV 23 10 upstream of Holstebro, a town with unacceptable in about 50% of Danish IV (seriously polluted) 82 35 30,000 inhabitants. Due to improved streams. Further improvements must All stations 651 651 methods of treating wastewater, P. therefore be achieved by the use of microcephalus now also inhabits the 27 other measures. First, the habitat qual- Table 10.2 Water quality, as measured km stretch downstream of the town of ity must be improved, thereby creating by the Viborg Index, for a number of Holstebro to the outlet in Nissum Fjord more variable and diverse substrate investigated stream sites in the County of [22]. In 2000 and 2001, P. microcephalus conditions. This can be obtained by Southern Jutland in 1986 and 1995–1997. was also recorded from two streams use of restoration measures and less The comparison shows that the macroinver- in another part of Jutland, where this intense stream maintenance. Second, tebrate fauna has become less impacted in stonefl y had never previously been the areas bordering the streams need the 10-year period. observed. to be managed as protective buffer In conclusion, the ecological qual- zones against inputs of sand and other ity of Danish streams has generally particulate material from surrounding improved and we have therefore fi elds. Third, an improved hydrological in the County of Southern Jutland with achieved better environmental quality interplay between stream and stream an increase in the number of slightly- for the money invested. Many sensitive valley allowing natural processes to impacted streams. species are now so widespread that occur and natural plant communities A change in the distribution of sen- they were removed from the Danish to develop is required e.g. by allow- sitive species is an alternative method Red Data Book when it was revised ing streams to periodically fl ood their of elucidating the general changes in in 1997 [2]. In recent years investiga- surroundings. Finally the improved the ecological quality. The County of tions in the streams have been more treatment of wastewater from scattered Funen has previously documented comprehensive and with more detailed dwellings and closure of storm water improvements in the streams [20] information of the distribution of outfalls would also enhance the general using the general guidelines from the individual species as a result. Therefore ecological quality. Ministry of Agriculture. These results the improved conservation status of Such concerted efforts are expected have later been supported by following many species may not only refl ect an to increase the proportion of streams changes in distribution of the mayfl y improvement in ecological quality, but where quality objectives are fulfi lled Heptagenia sulphurea and the stonefl y also a more comprehensive sampling from about 45% to 70–80%. Therefore, Leuctra fusca. The range of these two regime and a consequent increased we anticipate that there will be better species, both of which demand a rela- chance of fi nding sporadically occur- times for the macroinvertebrates in tively good water quality, has increased ring species. the future. We are on the way. The over the past 10 years [21]. A similar Danish counties and municipalities pattern for the genera Heptagenia and The future ecological status have performed a number of necessary Leuctra (not identifi ed to species) has Can we achieve optimal conditions measures, with more to follow. But it been found in other Danish coun- (fauna class 7) in all Danish streams? will take some years before wastewater ties. Furthermore many species that The answer is no. The prerequisite from scattered dwellings is treated and were previously restricted to head- for reaching water and habitat of an before stream habitat quality is suf- water areas; are now found further acceptable quality is often not avail- fi ciently improved. Therefore, we must downstream. These species are also able. But there remains scope for arm ourselves with patience. found in a number of neighbouring improvement in most streams. 122 RUNNING WATERS

Brede Stream was re-meandered during the years 1991–1998. 123

11 A new development: Stream restoration Streams have always been inﬂ uenced by man. Thousands of kilometres of streams have been straightened – often to the detriment of animals and plants. Heavy-handed maintenance has further deteriorated the conditions. In the 1990s, Denmark started a new approach. More environment-friendly stream maintenance methods were introduced and streams were being restored. What are the future perspectives?

Streams destroyed much physical variation as possible. The many thousands of kilometres The more diverse the environmental of streams in Denmark should be conditions in streams, the more diverse expected to sustain many diverse animal and plant life they can support. habitats with an animal and plant life If streams lack the natural variation rich in species. However, the variety of created by the shift between riffl es and animal and plant species in the Danish pools, animals and plants will have streams has declined dramatically fewer and less suitable habitats. throughout the last century. In 1864 Denmark had to surren- This decline is partly due to pollu- der almost a quarter of its territory to tion of the streams with wastewater Germany. This resulted in a shortage from agriculture, cities and industry. of cultivatable land forcing agricul- Streams were transformed into sewers ture to include former unexploited and diversion canals and were thus natural land and introduce more effec- prevented from sustaining any life. tive farming methods. Government However, introduction of the fi rst interventions and subsidies further environmental plans in the 1980s accelerated this development between has gradually solved the problem of the 1930s and 1960s. Many thousands wastewater, and the quality of stream of hectares of wetlands were drained water has been improved year by year. and thousands of kilometres of streams The stream water is once again so clean were straightened and deepened to that it should be able to fulfi l the goal quickly drain the water from the fi elds of large and varied populations of to the sea. animals and plants. The streams lost their natural shape, However, this goal will not be and regular dredging and very heavy- reached solely by stream water handed maintenance further deterio- becoming clean. In addition, streams rated the poor physical condition of themselves also need to contain as the streams. Because of their uniform 124 RUNNING WATERS

Figure 11.1 In 1992, there were 27 Salling Catchment Sound Obstruction obstructions in the 38 km2 catchment of Rødding Stream [1].

Rødding

Following the construction of the Tangeværk hydroelectric power station in the upper reach of the River Gudenå between 1918 and 1km 1920, salmon were unable to reach their spawning grounds. Only a few years later, salmon were extinct in this river.

Table 11.1 Reasons nature, such straightened and deep- Cause of problem Number of % ened streams only provide very poor for non-compliance stations conditions for freshwater life. with the stream Habitat degradation due to channelisation 101 38 quality objectives in The many dams and obstructions Hard-handed maintenance 35 13 constructed in connection with the three municipalities Sewage from rural areas 65 25 channelisation of the streams (see in Vejle County in Sewage from wastewater plants 27 10 Chapter 1), cause additional problems 1992 [2]. for the animals living in the streams. Low water discharge 17 7 The obstructions often hinder the free Organic pollution from freshwater fi sh farms 10 4 passage of animals up and down the Loading from agriculture 9 3 streams with varying effects. Minor Total 264 100 obstructions only have a moderate effect on the fi sh population, while larger obstructions prevent the migra- Regulated streams and heavy- environmental problems in 1992 were tion of certain invertebrates and many handed maintenance of streams attributable to poor physical conditions fi sh. If the animals are thereby hin- contrast sharply with naturally mean- and heavy-handed maintenance (Table dered from reaching their breeding dering streams with great physical 11.1). grounds, they might face the risk of variation in discharge, depth, sedi- extinction. The most famous Danish ment, vegetation and stream banks. The road towards an improved example is the extinction of salmon in This physical variation offers many physical environment the River Gudenå following the con- and very diverse habitats for plants Improvement of physical stream struction of the Tangeværket, a hydro- and animals. diversity in compliance with the objec- power station built between 1918 and It is an enormous task to improve tives set up by the Danish counties 1920. The hydropower station totally the physical conditions of degraded can be achieved by introducing more prevented the salmon from reaching streams in such a way that the physi- environment-friendly maintenance the upstream spawning grounds and it cal quality of the stream corresponds methods and by carrying out different therefore became extinct within a few with the clean water now fl owing in types of stream restoration. On this years. the streams. The poor physical vari- basis, stream restoration as a means In spite of concentrated efforts since ation is still the reason why so many of improving the physical environ- the 1980s to remove the obstructions, streams do not continuously live up ment was introduced in the Danish many still remain – and even in small to the biological objectives set by the Watercourse Act from 1982. Active streams obstructions may still be found Danish counties. In three municipali- restoration of streams is a quick and very close to each other (Figure 11.1). ties in the County of Vejle, 50% of the direct way of achieving the required 11 A new development: Stream restoration 125

improvement of the streams, and, in some cases, stream restoration may be Type 1 Type 2 Type 3 the only method for improvement. Restored streams should preferably resemble the former natural conditions as much as possible and should only require a minimum of future maintenance. Take for example, the removal of all obstructions to Figure 11.2 Schematic outline of the three types of restoration project. Each of the three re-establish the free passage of fauna project types is described in the text. upstream and downstream. Where this is not possible, free passage should be re-established by a constructed or restored riffl e or a new bypass stream. Figure 11.3 Map of Denmark indicating the location of the three types of restoration Fish ladders, on the other hand, should projects. preferably be avoided as they are selective; being often only passable to strong swimmers, such as salmonids, Type 1 and hindering the passage of other Type 2 fi sh and invertebrates. In contrast, all Type 3 species of fi sh and stream invertebrates Stream are able to pass a correctly constructed bypass stream. In addition, riffl es and bypass streams can function as part of the natural stream. It is primarily the counties that have carried out the restoration projects. Whilst many counties and municipalities have initiated restoration projects they have often received fi nancial support from the National Forest and Nature Agency. The State, represented by the National Forest and Nature Agency, has also been responsible for a few projects – e.g. the River Skjern Project (see Box 11.1 on page 126–127). In co-operation with the counties, the National Environmental Research Institute has gathered and compiled information on many restoration projects in a database [7]. Projects have been divided into three types (Figure 11.2); those where restoration has been carried out locally on short stretches of the stream (type 1), those where restoration re-establishes the continuity 126 RUNNING WATERS

in the stream (type 2), and where the restoration includes both the stream 200 200 and its adjacent areas in the stream valley (type 3). 150 150 Type-1 projects typically result in local improvement of stream and 100 100 bank habitats. Type-2 projects restore free passage and continuity along the 50 50 Number of projects stream by reconnecting reaches and Number of type-2 projects their nearest surroundings. Fauna 0 0 are thereby enabled to migrate freely < 85 < 85 year year between different stream reaches and 85-86 87-88 89-90 91-92 93-94 95-96 97-98 85-86 87-88 89-90 91-92 93-94 95-96 97-98

between the stream and its close sur- Unknown Unknown roundings. Type-3 projects involve Type 3 (n = 34) Other methods (n = 92) both the stream and its valley. An Type 2 (n = 807) Obstruction removed (n = 533) Type 1 (n = 227) Bypass stream (n = 78) example would be a project seeking to Fish ladder (n = 104) raise groundwater levels in fl oodplain meadows and thus encouraging the Figure 11.4 The number of implemented Figure 11.5 Type-2 projects (split into four stream to fl ood at high water levels. restoration projects between 1985 and 2000. groups) that reconnect the stream reaches Higher water level and more frequent Notice the shift over the years in the type of and restore free passage and continuity of fl ooding are desirable when one wants projects being implemented. streams. to reduce the transport of sediments or the concentration of nitrogen or ochre in the stream. Restoration methods are Restoration methods have changed It has also become easier to award generally contrary to methods used in over time (Figure 11.5). At the time landowners one-off compensation in the past to drain the meadows. when the restoration of continuity the form of either money or new agri- By 1999, more than 1,000 restoration between stream reaches was initiated, cultural land. An increasing number projects had been carried out by the the most frequent method was to set of landowners now accept restoration Danish counties. These projects ranged up or build fi sh ladders. Gradually projects and many even warmly sup- from laying out spawning gravel to it became obvious that fi sh ladders port the idea of restoring the dynamics comprehensive projects re-meander- often only allow the passage of strong and variation in streams as much as ing streams and improving connectiv- swimmers like trout and furthermore possible. ity between the stream and its valley that they require intense maintenance. Re-meandering of the Brede Stream (Figure 11.3). Consequently, bypass streams that in the County of Southern Jutland The database informs about the go round obstructions, that cannot be proves the good intentions. This was division of the project types (Figure removed entirely, are now preferred. the second largest Danish river restora- 11.4). Among other things, it gives a These are much easier for fi sh and tion project, only surpassed by the historic view of the development of invertebrates to negotiate and do not River Skjern project. The Brede Stream stream restoration. The reconnection require as much maintenance. was re-meandered during the years of continuity between stream reaches 1991–1998 and this involved almost 100 (type-2), for example, was given espe- Restoration of the Brede Stream landowners. The once 19-km straight cially high priority during the 1990s, It remains diffi cult to convince Danish river was extended to a meandering while type-3 projects involving both landowners along streams to accept river of 26 km (Figure 11.6). Some the stream and its valley have gradu- a restoration project. However, with negotiation between the county and the ally received higher priority in recent steadily increasing experience of river fi rst landowners involved was required years. restoration, the situation has improved. to get the project running in 1991. But 11 A new development: Stream restoration 127

as the years went by and the remain- Monitoring was initiated one year ing landowners along the river saw the before the excavation and ran for the positive results of the re-meandering, subsequent two years. Among other and as the compensations in connec- things, it showed that there were tion with the re-meandering became less nutrients and ochre in the water more attractive, it became easier for two years after the re-meandering. the county to obtain the landowners’ The same also partly applied to the acceptance. The restoration was car- transportation of sand. If gravel banks ried out on a voluntary basis and all in the stream are infi lled by sand they requests from landowners, regarding become unsuitable as spawning banks the course of the river, were taken into for fi sh and as habitats for inverte- consideration. brates. The monitoring period proved When the re-meandering of the too short for an unambiguous evalu- river was initiated in 1991 it was, as ation of the effects of the re-meander- previously mentioned, one of the fi rst ing on animals and plants. However, projects of its kind to be carried out two years after the re-meandering the in Denmark. To estimate whether the number of species and individuals was project would meet the expectations higher or at least at the same level as of improved water quality, improved before re-meandering. physical conditions and greater biodi- Scenically, the Brede Stream has versity, a comprehensive monitoring become much more attractive than programme was carried out in connec- before and it is now a popular recreation with the second stage initiated in tional area for both anglers and the 1994. local people.

Fish ladders (top picture) are often only passable for species that are strong swimmers. Bypass streams, on the other hand, extend the water drop over a longer reach and are therefore more often passable to all species of ﬁ sh and invertebrates. The lower picture shows the 655 m-long bypass stream at the Holstebro hydroelectric power station on the River Storå.

The old straight run of the Brede Stream is being re-ﬁ lled after the opening of a new meander. 128 RUNNING WATERS

Figure 11.6 Brede Year 1991 1994 1995 1996 1997 1998 Total Stream was previously a straight Number of landowners 8 22 16 17 19 15 97 river of 19 km. After Length before (km) 2.7 2.8 3.6 4.0 3.4 2.8 19.3 1998 restoration between Length after (km) 3.2 4.3 4.3 5.4 4.8 3.7 25.7 1991 and 1998, 3 1997 Earth excavated (m ) 45,000 72,000 42,000 50,000 44,000 25,500 278,500 it now meanders Area of the river valley (ha) 40 60 70 80 65 60 375 along a stretch of Costs excluding taxes 240 470 415 335 325 335 2,120 26 km. (thousand EURO) 1996

Bredebro

Before remeandering 1995 1km Løgumkloster After remeandering 1994 1991

Maintenance as a tool for restoration cheap method and should therefore be about by environment-friendly mainte- It is not only by restoration that we can used to a greater extent in connection nance take place gradually over longer change the physical conditions and the with the rehabilitation of degraded periods. shape of streams. Benefi cial changes streams and wetlands. In addition, if An excavator may dig new mean- can also be obtained if routine mainte- regular maintenance is stopped, the ders. Piped streams may be re-opened, nance of streams is modifi ed or totally plant community in the streams will obstructions removed, and new gravel stopped. Streams have traditionally become richer and more diverse [10]. and stone bars established in previ- been maintained in order to quickly The areas adjacent to the stream ously sandy streams. However, it is drain water from the fi elds. This has will also be fl ooded more frequently, usually necessary to use excavators been achieved by cutting all vegetation which, on one hand, may interfere to achieve an actual restoration, and several times per year and by removing with the agriculture, but on the other they are expensive to run. An estimate gravel and stones. In that way streams hand may have important benefi ts for of the total length of Danish streams have been kept in their unnatural society as a whole. To some extent, the improved during ten years of digging canal-like course to the detriment of public demand for reduced nitrogen new meanders or re-opening piped animals and plants. discharges to fjords and the sea will stretches shows that it is less than 100 With limited or no maintenance, be met by the natural stream valleys kilometres – and maybe only half of the physical conditions of streams will retaining nitrogen by means of more that. As a re-meandering project costs revert to a more natural state. Vegeta- frequent fl ooding. The saved costs EURO 70,000–150,000 per kilometre of tion functions as a biological engineer from not having to maintain streams stream, it is totally unrealistic to have in streams and may, if left undisturbed, several times a year to keep them in all streams re-meandered by active res- change physical conditions [8]. In- their present canalised and unstable toration. Changed maintenance meth- stream and bankside vegetation will state should also be considered in the ods, on the other hand, may improve relatively quickly contribute to narrow- overall evaluation. the physical condition of thousands ing the profi le of the stream, advance However, changed maintenance of kilometres of streams at a far lower sediment deposition, and strabilise the methods do not attract the same public cost. So, in spite of the longer time stream bed [9]. In time, this will result attention as restorations. With resto- scale, it is the sympathetic maintenance in a more naturally meandering and rations, the public can see improve- that will promote good physical condi- elevated stream. A change in stream ment in the stream from day to day. In tions in the majority of streams. maintenance regime is a relatively contrast, the enhancements brought 11 A new development: Stream restoration 129

Weed cutting with a scythe is only possible in shallow streams. Although some vegetation is retained in the stream after cutting, the water does not necessarily ﬂ ow slowly. Water crowfoot has the majority of its biomass at the water surface, which allows the free passage of water below the water surface.

What about the future? lishing all experiences, good as well Danish counties and the Ministry of as bad. Each project should include an Environment generally prioritise river element of monitoring, and adequate restoration highly as demonstrated resources should be allocated for more by the resources granted since the comprehensive scientifi c monitor- introduction of Danish river restora- ing programmes to investigate which tion. More than EURO 15 million has stream restoration methods are the been spent on 550 of the more than most effi cient and cost-effective for 1,000 projects carried out so far by the achieving improved biodiversity. counties. In spite of this, restoration It is also important to allocate some projects are infrequently followed up funds for the monitoring of the long- by monitoring of the biological, physi- term effects of restorations. Too many cal and chemical effects on streams and of the monitoring programmes only Only small forested streams are today in a their adjacent areas. If the effectiveness describe conditions just before and pristine state. All other Danish streams need of a project is monitored, far too often after the project was completed. This to be restored to improve their quality. the experiences and conclusions are not applies, for example, to the monitor- published for the benefi t of others, but ing of the Brede Stream. Short-term instead are prepared solely as internal monitoring cannot contribute to a fi nal reports. Consequently, in this chapter evaluation of the biological, physical we are not able to take a closer look at and chemical effects. As restoration the possible effects of the large restora- projects often involve a pronounced tion projects on the animal and plant interference in the stream, it is impor- life in streams. There is therefore a tant to follow the long-term effects. great need to examine the effects of dif- This will generate invaluable knowl- ferent restoration approaches and the edge on how future restoration projects methods used. This could be achieved should be carried out and how they by setting aside a certain percentage may be prioritised to obtain rich and of restoration grants for subsequent varied plant and animal communities monitoring programmes, and by pub- in our streams. 130 RUNNING WATERS

Box 11.1 River Skjern – the most comprehensive Danish river restoration project

and large amounts of hay could be saved. However, the river was also unpredictable. Large ﬂ ooding could occur at any time of the year and thereby destroy the haymaking – or the whole harvest of the year could disappear into Ringkøbing Fjord. For centuries the locals attempted to protect the meadows and control the river water by regulation. As early as 1901–02, the main channel of the River Skjern from Lønborg town to the fjord was partly regulated and secured with summer dikes. But even after this “the great river regulation”, there were many wet and soft-bottomed areas where it was impossible to move about with horses and machinery during summer. And during winter the river continued to ﬂ ood the area.

After World War II, a large number of drainage projects “But soon the areas will become a major attraction. As the were carried out throughout Denmark with the support of dikes become covered with grass and the roads become pass- the State and the public in order to obtain agricultural land able, when it will be possible to wander along the river for for food production. Also the landowners in the River Skjern hours, when the sand banks have gone and boats are able to valley wanted new land for grain growing. Following numer- sail from Borris to the fjord – well, then we will see crowds of ous disputes the landowners fi nally obtained governmental tourists that will even surpass the number of 50,000 that each support, and in 1962, Denmark’s largest drainage project could year visit the drained Fiil Lake. And it will become a publicly begin. Over six years, 4,000 hectares of meadow and marsh- recognised nature experience.” land were transformed into fi elds. The project involved the Representative of the landowners, farm owner J. Sme- embankment of the downstream part of the river and its large degaard-Mortensen, at the beginning of the drainage in tributaries. The groundwater level was lowered and fl ooding 1963 [3]. was hindered by drainage canals and pumping stations. The regulation was completed in 1968, but the fi rst oat crop from “The meanders and natural variation in water level will the new fi elds was harvested as early as in 1966. The drainage contribute to improving the general habitats for animals and project totalled DKK 30 million corresponding to DKK 220 mil- plants, it will result in improved river quality for the River lion (or EURO 30 million) in 1999 prices. The State covered 70% Skjern catchment and Ringkøbing Fjord. And the recreational of the expenses. and tourist value of the area will be many times multiplied.” Minister of Environment and Energy, Svend Auken, at There is hardly any doubt that the drainage project proved the beginning of the restoration in 1999 [4]. to be a fi nancial success for the farmers. On the other hand, the project undoubtedly also had negative consequences for The River Skjern is Denmark’s largest river in terms of dis- nature and the environment of the river, its surroundings and charge. It has an average discharge of 35 m3/s. Upstream of Ringkøbing Fjord. Plant and animal habitats have deteriorated, Borris town the river runs in its natural meandering bed for and the increased load of nutrients and ochre have impover- long stretches, while it has been totally regulated and dikes ished conditions for life in Ringkøbing Fjord. At the same time, built along its length from Borris town to Ringkøbing Fjord. 33 years of drainage and cultivation have caused the peaty soils to sink by up to one and a half metres. If the agricultural For centuries, the River Skjern valley has played an impor- production was to continue, it would have required a new tant role for life and agriculture in the poor soils of Western drainage project in the near future. Jutland. The river fl ooded its adjacent areas during winter and spring, thereby depositing ‘“fertiliser” as nutrients and In the light of such environmental deterioration, a large major- organic matter. Consequently, the meadows were productive ity in the Danish Parliament decided in 1987 that a nature 11 A new development: Stream restoration 131

rehabilitation project should be carried out in the River Skjern The total cost of the project is estimated at about DKK 250 valley. In 1990, the State started buying and re-allocating million (about EURO 35 million). This is very similar to the cost land. In 1997, the Danish Forest and Nature Agency published of the drainage project when it was carried out in the 1960s, if a detailed restoration proposal, and in 1998 the Parliament calculated in prices of the time. passed the project construction act. It included 2,200 hectares of the river valley from Borris to Ringkøbing Fjord and A total of EURO 1.2 million has been allocated for a project increased the river length from 19 km to 26 km. A 3 km regu- monitoring programme which will include the following main lated stretch of the Omme Stream, a major tributary to the elements: River Skjern, was also included in the project and increased the river to 5 km of meandering river. The new meandering river • Hydraulics, stream morphology, groundwater, and frequency course primarily followed the original meanders present before and extent of fl ooding the initial comprehensive engineering of the early 1900s. • Water and material transportation, water quality and sedimentation The purpose of nature rehabilitation was to re-create the • Vegetation in the river and its adjacent areas extensive natural areas of former times and to improve the • Aquatic invertebrates habitats for animals and plants. For the benefi t of Ringkøbing • Fish – especially the development of salmon Fjord, the project was also seeking to partly restore the self- • Birds and otter. purifying ability of the river valley to clean the water during the regular fl ooding. Re-meandering of the river is expected to result in a more varied animal and plant life in the area. Especially plant In June 1999 the Minister of Environment initiated the larg- species such as the rare River Crowfoot and Water Plantain, est river restoration project ever undertaken in Denmark. The and animal species such as salmon and otter are expected to project had three phases and was to be completely fi nished in regain ideal habitats. The River Skjern valley is also expected autumn 2002. The fi rst phase, from Highway 28 to the Fjord, to become a more attractive area for many nesting and resting was fi nished in October 2000, when the water was released into birds. the new meandering course. The second phase, from the inlet of the Omme Stream to Highway 28, was fi nished one year later. The project is further described in [4] and [5]. Finally, the project was completed in autumn 2002, when the water was diverted into the fi nal part of the restored course Maps illustrating the downstream area of the former and future from the town of Borris to the Omme Stream in autumn 2002. run of the River Skjern and the Omme Stream [6].

BORRIS SKJERN Area included in the restoration project Wet areas New restored run River Old regulated run Skjern

Ganer Stream Highway 28 River Skjern

N Omme Stream

Tarm Brook

Tarm Møllebæk Gundesbøl Stream Brook 132 RUNNING WATERS

Danish streams are exposed to many different pressures. 133

12 Environmental state and research Even though the environmental state of Danish streams has generally improved in recent years, for various reasons less than half of them meet their current environmental quality objective. The Water Framework Directive, recently adopted within the EU, will no doubt inﬂ uence future management of Danish streams. This chapter summarises past developments in Danish stream research and describes the areas where our knowledge of the biological aspects needs to be enhanced in the coming years to ensure that the stream authorities have the necessary tools to improve the country’s streams.

Streams in Denmark streams. As the EU Water Framework There are more than 60,000 km of Directive will unavoidably infl uence natural and man-made streams in future Danish stream management, the Denmark, i.e. more than 1 km of stream Directive’s intentions will be briefl y per km2 of land. Streams are therefore reviewed. important elements in the Danish The following account emphasises landscape and are of considerable the ecological quality of the Danish importance to the biodiversity of fl ora streams, focusing on their biologi- and fauna. cal state. However, the streams are At the same time, though, human also important with regards to input, population density is more than 100 transport and turnover of nutrients persons per km2 and about two thirds and hazardous substances derived of Denmark’s area is exploited for from households, industry, aquacul- intensive agriculture. Streams are con- ture and agriculture. These processes sequently one of the ecosystem types are of great signifi cance, not least for under greatest anthropogenic pressure. downstream lakes and marine waters. This imposes great demands on stream Our present knowledge of a number management. of these aspects is insuffi cient and One of the tenets of Danish nature concerted efforts are therefore needed and environment policy is that it must to strengthen knowledge in these areas, be fi rmly rooted in sound knowledge. in order, among other things, to be able The aim of this chapter is thus to iden- to model the processes. For the sake of tify some of the most urgent strategic brevity these aspects will not be dealt research needs on the basis of an analy- with here, however. sis of the environmental state of Danish 134 RUNNING WATERS

Environmental state of Danish streams In Denmark, as in most of Europe, the environmental state of streams is Wastewater used to traditionally assessed from the compo- be the main cause sition of the macroinvertebrate fauna. of the unsatisfac- The method currently employed is the tory environmental Danish Stream Fauna Index (DSFI, see state of Danish page 118) [1] whereby an index value streams. on a scale from 1 to 7 is calculated from the composition of the stream fauna. Pristine streams containing Danish streams [3]. More of the streams factory environmental state of 3,000 many sensitive species are assigned in Jutland (50%) fulfi lled their objec- km out of the 11,600 km of stream. an index value of 7, while the most tive than those on the island part of It is important to note, though, that heavily disturbed streams are assigned Denmark (30%). concerted efforts to reduce wastewater an index value of 1. Due to the natural Since the 1970s, biological assess- discharges from sparsely built-up areas physical conditions pertaining, some ment of streams has been undertaken will not necessarily result in the quality of the streams will score less than 7 many thousands of times in connection objective subsequently being met in despite being completely unaffected by with routine stream supervision by a further approx. 3,000 km of stream. anthropogenic factors. the counties. There is little doubt that While this measure is necessary, it is In 1997, the nationwide programme the environmental state of the Danish not necessarily suffi cient – for exam- to monitor the aquatic environment streams has generally improved. The ple because of unsatisfactory physical encompassed 444 representative stream distribution of a number of clean-water conditions, discharges of pesticides and stations. An analysis of the results species has increased markedly on ochre, etc. shows that the DSFI was 5–7 in 37% Funen and in Jutland, several species of the streams, 4 in 43% of the streams have been removed from the Danish The EU Water Framework Directive and even lower in the remaining 20% Red List [4], and the trout population The overall objective of the Directive is [2]. The environmental state is gener- has increased considerably [5]. Since to improve the environmental quality ally better in large (>2 m wide) streams a suffi ciently standardised procedure of European water bodies wherever than in the smaller streams. Due to the such as the DSFI was not previously they are, at present, unsatisfactory, as natural physical conditions pertaining, used in stream supervision, it is not well as to prevent further deterioration. the environmental state of streams in possible to quantify the improvement The objective is to be attained through Jutland is generally better than on the on the basis of the faunal class. the preparation of action plans for island part of Denmark. whole river basins and by document- Each county’s Regional Plan stipu- Pressures affecting stream ing possible improvements through lates environmental quality objectives environmental state monitoring. for their streams. Three main categories The counties have analysed why the Ecological quality is one of the key of objectives are employed: stringent quality objectives are not met in 11,600 phrases, and it is to be documented objectives, basic objectives and eased km of stream [3]. Among the many dif- using physical, chemical and biological objectives. Streams with stringent ferent factors to which this is attribut- indicators. The emphasis is expected objectives must be completely unaf- able, the most important is wastewater to be on the biological indicators. For fected, whereas those with basic objec- from sparsely built-up areas followed streams these will include not only tives may be slightly affected and those by stream regulation/maintenance. macroinvertebrates, but also mac- with eased objectives may be severely The counties thus estimate that rophytes and fi sh. Indicators for the affected. In the period 1993–1996, qual- wastewater from sparsely built-up quality of the riparian areas will also ity objectives were only met in 45% of areas is the main reason for the unsatis- be included due to the tight coupling 12 Environmental state and research 135

• Enhanced recognition of the need to consider streams as an integral part of the landscape and focus on the interplay between streams and the riparian areas • Enhanced international coopera- There is a tradition of tion on the build-up of knowledge stream ecology research in related to stream management Denmark. Research on the ecological interaction between the surroundings and between streams and riparian areas. the National Environmental Research the stream and the mutual interaction Another challenge in the Directive is Institute) was established in Silkeborg between plants, animals and fi sh has to create a system able to ensure that together with the Freshwater Fishery not previously been accorded high good ecological quality in, for exam- Laboratory (now the Danish Institute priority in either Denmark or at the ple, an Italian stream is comparable of Fisheries Research). This markedly European level. The Water Frame- with good ecological quality in, for strengthened research aimed at sup- work Directive is likely to help change example, a Danish stream. Important porting the management of Danish this, however. Enhanced cooperation tools for ensuring this will be methods streams. between Danish researchers and our for determining the baseline state, i.e. European colleagues is a prerequisite if the ecological quality in a completely Present structure of Danish stream this is to succeed. pristine stream, as well as methods to research measure deviations from this state. Danish stream research is currently Perspectives for Danish stream distributed among a number of uni- research History of stream research in versities and sector research institutes. In this section, we describe areas where Denmark The research among the universities there is a need for concerted efforts Stream research has been conducted is chiefl y conducted at the University to support the future management in Denmark for more than a century. of Copenhagen Freshwater Biologi- of Danish and European streams, For many years research was predomi- cal Laboratory, while that among the i.e. those pertaining to strategic and nantly concerned with the fl ora, fauna sector research institutes is mainly car- applied research. or natural history of the streams. The ried out by the National Environmental main actors were scientists such as C. Research Institute and the Danish Insti- Stream quality objectives Wesenberg-Lund, P. Esben-Petersen, tute of Fisheries Research in Silkeborg. Implementation of the EU Water H. Ussing and C. V. Otterstrøm. The In recent years some positive trends Framework Directive necessitates fi rst major quantitative description of have been detectable in Danish stream the establishment of comparable and a Danish river system was the Suså research of both an organisational and quantifi able European-wide indicators Stream surveys carried out in the 1940s a cognitive character: of the ecological quality of streams under the direction of Kaj Berg [6, 7]. and their riparian areas. The relevant As early as 1925, C. Wesenberg- • Enhanced cooperation between indicators for Danish streams are mac- Lund drew attention to the serious basic research at the universities and roinvertebrates, macrophytes and fi sh. pollution of our lakes and streams [8]. the strategic/applied research at the While there is a century of experience However, as late as 1975, few pol- sector research institutes with macroinvertebrates, considerable lution-related studies had still been • Enhanced cooperation between fi sh- work is needed before operational mac- conducted, [9–11]. In 1976, the Danish eries research and environmental rophyte and fi sh-based indicators can Environmental Protection Agency’s research be developed. Freshwater Laboratory (now part of 136 RUNNING WATERS

In order to be able to establish rea- other means, through the establishment ecosystems meet. Improved knowledge sonable quality objectives, the natural of spawning grounds and habitats for of the interaction between streams and state of the stream needs to be known, fi sh. It is striking, though, that the envi- riparian areas can help ensure the suc- i.e. the biological communities that ronmental effects of the various meas- cess of initiatives aimed at improving would be present under the natural ures are very poorly documented both the quality of nature in general. physical and chemical conditions. Pre- nationally and internationally. At best dictive modelling is a promising tool the effects on the macroinvertebrates Fishery and aquaculture for predicting the probable composi- and/or fi sh have been recorded, while Recreational fi shing is a characteristic tion of the fl ora and fauna and hence investigations of the plant communities feature of Danish streams. Fisheries the ecological quality of streams on and the long-term effects at the system research in Danish rivers has chiefl y the basis of information on the physi- level are virtually non-existent. focused on species such as trout and cal and chemical conditions [12, 13]. To improve this situation, the salmon and has often concerned opti- Such a tool needs to be developed for European Centre for River Restoration misation of these populations through macroinvertebrates, macrophytes and (ECRR) was established in 1994 founded stocking. In recent years, genetic stud- fi sh in order to be able to comply with by EU LIFE. The secretariat was at ies have been initiated to preserve the the intentions of the Water Framework NERI until 2002, and then moved to natural fi sh populations and genetic Directive. In Denmark too, a system is RIZA, the Netherlands. Through variation. needed for establishing quality objec- conferences, a web site (www.ecrr.org) Fisheries research and environmen- tives on the basis of objective criteria, and other forms of networking the tal research have traditionally been especially regarding the physical con- ECRR promotes river restoration as a separate disciplines. Cooperation has ditions. Such a system would ensure cost effi cient tool for improving river now been initiated to investigate the that stream quality objectives are quality and stimulates river restoration signifi cance and dependence of fi sh in established in a comparable manner research and dissemination of the a broader ecosystem context. Some of throughout the country. fi ndings. the important themes are: the interactions between fi sh and the other bio- Stream maintenance and restoration Interaction between streams and their logical communities, fi sh as indicators In order to ensure rapid drainage riparian areas of ecological quality and methods for of agricultural land, the majority of At both national and international maintaining and restoring streams so Danish streams have been channelised. levels, considerable research has been as to ensure a rich and varied fl ora and The spatial variation in current velocity conducted on the interaction between fauna, including fi sh. and substratum conditions has conse- forests and streams e.g. the input and Aquaculture in the form of fresh- quently been reduced in the reaches biological processing of leaves and the water fi sh farms is a major reason for in question, leading to deterioration structure and role of the macroinver- the failure of many streams in Jutland in habitat conditions for plants and tebrate community. However, as most to meet their quality objective. In 1997, animals. Stream maintenance through Danish streams lie in open landscapes, the 433 freshwater fi sh farms in opera- weed clearance and sediment removal it is likely that the structure of the tion together produced approx. 32,000 further reinforces the uniformity. riparian areas is of considerable signifi - tonnes of rainbow trout [14], approx. In Denmark, since the mid 1980s, cance for their biological communities. 95% of which were exported, repre- stream maintenance practice has For example, the structure of riparian senting an annual turnover of DKK 482 changed markedly in the direction of areas can infl uence the dynamics of the million (65–70 million EURO) [15]. The environmentally sound maintenance bank-side plant communities which in fi sh farms affect streams through dams or no maintenance at all [3]. More over, turn can infl uence the survival of adult interrupting the streams’ natural conti- more than 900 restoration projects insects and hence the possibilities for nuity, through the discharge of organic have been carried out in large Danish macroinvertebrate recruitment. Plant matter, nutrients, medicines and other streams in which the physical condi- and animal diversity is particularly ancillary substances often leading to tions have been improved, amongst great in places where different types of “dead” stream reaches alongside the 12 Environmental state and research 137

fi sh farms [16]. If fi sh farming is to be sustainable, research and development work will be needed to ensure that production is compatible with fulfi l- ment of the streams’ quality objective. Cross-disciplinary research in this area Laboratory identi- has recently been initiated. fi cation of insects and extraction of Hazardous substances chlorophyll. An estimated 100,000 different products containing up to 20,000 different chemical substances are estimated to biological communities and their inter- Fortunately, there is a strong tradition be on the market in Denmark. All these action and to further identify which of close contact between researchers can potentially end up in the aquatic substances affect the streams and how. and environmental administrators in environment and hence affect its envi- Denmark. This is fostered through ronmental state. Research and the future scientifi c meetings aimed at keeping Little is presently known about the As is apparent from the above, Danish the administrators abreast of the latest occurrence and distribution of haz- stream research faces many challenges. research fi ndings and the researchers ardous substances in Danish streams. In the 1970s it was believed that, once abreast of the unsolved questions of The Danish Aquatic Monitoring and treatment of urban wastewater was importance to the county and munici- Assessment Programme NOVA 2003 implemented, the quality objectives pal environmental administrators. It is – and after 2004, NOVANA – has would be attained. When treatment of the responsibility of the politicians to provided important information on the wastewater had been implemented, decide what the environmental state approx. 300 substances, especially with however, it became obvious that poor of Danish streams should be but it is a regard to point sources. Pesticides are physical conditions often hindered joint responsibility of the administra- used in large amounts, and it is known attainment of a satisfactory stream tors and researchers to provide the that they are detectable in streams. In quality. We know that the future will politicians with a sound basis for deci- Funen County, pesticide contamination also hold surprises in store, but not sion making that is both relevant and is purportedly the reason why 15% of which ones. founded on the latest knowledge. the stream reaches fail to meet their Some of the areas where research is It is impossible to stipulate stringent quality objective. Little is known about needed in order to be able to improve criteria for what is the most relevant the impacts of pesticides on the stream the environmental state of streams research. Less than half of Danish fl ora and fauna, though. These could in the long term have been outlined streams meet the politically determined involve direct toxic effects, although it above. Common to the majority of quality objectives and the reasons for is more likely that indirect effects are these is an enhanced understanding of this depressing fi gure are diffi cult to involved whereby the substances affect the physical-chemical- biological inter- determine. If research efforts in the the interaction between the various actions within streams and between areas outlined above can help signifi - organisms. streams and their riparian areas. cantly enhance the number of streams Some of the other substances Broad-based research in clearly defi ned meeting their quality objective, then discharged into streams from point areas is a prerequisite for procuring the the research will have been relevant. sources undoubtedly have similar knowledge and tools needed for stream effects, but these are as yet unknown. management. In the coming years it is therefore Interplay between research and necessary to develop experimental county/ municipal stream manage- methods for investigating the effects ment is crucial if the research fi nd- of various chemical substances on the ings are to be utilised in practice. 138 RUNNING WATERS

Streams are stimulating and adventurous. 139

13 Streams and their future inhabitants In this ﬁ nal chapter we look ahead and address four questions: • How do we improve stream management? • What are the likely developments in the biological quality of streams? • In which areas is knowledge on stream ecology insufﬁ cient? • What can streams offer children of today and adults of tomorrow?

Improved management have not been kept free from other pol- When you are placed at a conven- luting or disturbing impacts deriving ient distance from the legislators in from freshwater fi sh farms, agriculture, the Danish Parliament and the local forestry, market gardens, sewage from politicians, it is easy to wonder at scattered dwellings and water abstrac- the apparent lack of effectiveness in tion. There has been a lack in either the management of streams, and the political determination or suffi cient distribution of competence and power administrative authority. between the state, the counties and the The required treatment of domes- municipalities. Let us take a look at tic sewage has been implemented two conspicuous conditions. throughout Denmark, while destructive dredging of the stream and Consequent or realistic planning? washing down of ochre from the land Thanks to heavy investments in drains have continued. In many forest biological and chemical purifi cation of springbrooks, where there are excel- domestic wastewater, oxygen deple- lent opportunities of achieving a rich, tion and smothering of the stream bed almost unaffected fauna, this high qual- is no longer found in most streams ity status has not been reached because that were formerly heavily polluted. In destructive impacts such as abstraction consequence, the biological quality of of water for freshwater fi sh farms and streams based on the composition and domestic water supply, or drainage for richness of large invertebrate species the enhancement of forestry produc- has improved (page 121). Unfortu- tion have been tolerated. In the vicinity nately, there has been an increase in of Copenhagen, the groundwater the number of streams of intermediary table has been drawn down by 5–13 m quality, which are neither very clean and numerous springs have dried nor very polluted. In many cases the up. In addition to the direct impacts small percentage of very clean streams on streams and their hydrology, the 140 RUNNING WATERS

terrestrial vegetation has undergone profound changes. Mixed deciduous forests have been replaced by planted There has been monocultures of conifers along some a serious decline of the most pristine streams. Because in Potamogeton of the intimate coupling between land praelongus (long- and water, such changes have had a stalked pondweed) severe negative effect on the fl ora and in NW Europe. fauna of the stream, such as acidifi cation, enhanced shading and a poorer nutrient quality of coniferous leaf litter special permits. It is a paradox that the to disperse passively with the current for microbial decomposition and con- high physical and biological quality of from populations that have survived sumption by invertebrates. streams restored by means of goodwill the disturbance. Fish can swim actively and the large investments of some to all reaches in the stream system, and Zig-Zag course in management municipalities is actively destroyed in insects can disperse both along and The municipalities are responsible for other areas by contrasting actions of between the streams as fl ying adults. the management of small local and people with different priorities. Moreover, plants and animals produce privately owned streams, and they keep numerous offspring that may disperse control on the sources of pollution (e.g. Development of plants, invertebrates with the current and large animal agriculture and market gardens). The and fi sh vectors (i.e. fi sh, birds and humans), municipalities are usually small admin- Stream plants and animals are gener- giving the organisms the potential for istrative units with limited professional ally tolerant of sporadic disturbances rapid colonisation of depleted reaches. expertise. Environmental issues are and are able to re-establish former It is an essential prerequisite for often treated with a rather mixed degree communities [1]. Following a pollution rapid re-establishment of lost popula- of attention and understanding. Some incident which kills most invertebrates tions that invertebrates and fi sh must municipalities make a great effort to and fi sh along a stream reach, or after have survived in parts of the stream secure the biological quality of streams, extensive weed cutting and dredg- system, and preferably in close proxim- often encouraged by strong environ- ing, the communities usually become ity to the affected sections (page 80). mental awareness of the public. The re-established within a relatively short Re-establishment of plants is most municipalities may even undertake res- period. Depending on the circum- successful when populations still grow toration projects such as remeandering stances and the type of organism, upstream of the affected reaches (page of streams and re-establishment of stone the recovery may take a few weeks, 49). and gravel beds that have been lost months or 1–2 years [1]. There are Where species have not survived in during decades of intensive dredging. several reasons for this overall robust- refuges in the stream systems, recolo- In other municipalities with strong ness. Streams are naturally exposed to nisation will be slow even though the agricultural interests, however, the atti- substantial fl uctuations in discharge environmental conditions may become tude and administrative practice may and fl ow velocity leading to erosion suitable. Within a few years or decades be quite the contrary. Small streams of the stream bed and a catastrophic species can recolonise from neighbour- with very clean water are physically loss of organisms. Stream species have ing stream systems, but successful re- destroyed by severe dredging. Huge therefore evolved the ability to cope establishment becomes more and more machinery is used to excavate the with disturbances either by avoiding diffi cult as the distance to neighbour- streams to a greater width and depth them, taking shelter, or resisting the ing populations increases and stronger than strictly necessary to transport disturbances by strong anchoring and barriers have to be passed. The likeli- the water. Stones are left on the banks reduction of drag forces by changes in hood of recolonisation is reduced even and the entire operation resembles shape and orientation. Streams are con- further if the potential donor popula- a development that usually requires nected systems allowing all organisms tions are few and small. 13 Streams and their future inhabitants 141

Will rare plants return? [2] . Among the 25 known species of Extinct Threatened/ All Within the large plant genus Potamoge- stonefl y, the two large, impressive spe- species rare species species ton several species which were wide- cies Perlodes dispar and Dinocras cepha- Damselfl ies 4 17 50 spread and locally abundant are now lotes have disappeared from Denmark. Mayfl ies 5 15 42 restricted to very few localities and From 1980 to 2000, several species Stonefl ies 2 8 25 have small populations (page 109). In have become more widespread and Caddisfl ies 10 44 167 general most aquatic plants are widely abundant (page 121). Among may- Blackfl ies 0 7 24 distributed regionally as well as glo- fl ies this expansion includes Ephem- bally. This widespread distribution can era danica, Heptagenia fuscogrisea, H. Table 13.1 Many species of Danish be explained by the high probability of sulphurea and Paraleptophlebia submar- freshwater insects are threatened or have propagule establishment because open ginata which are no longer considered become extinct. space frequently becomes available as threatened species. Among stone- on the stream bed. Dispersal depends, fl ies, Isoptena serricornis has virtually nevertheless, on the existence of viable disappeared in Europe, but the species populations. In that respect the premise have become more abundant in some of readily available propagules has Danish rivers (the Varde Stream, changed dramatically in Denmark Karup Stream and River Skjern). Also during the past 100 years. Perlodes microcephalus has become more Many pondweed species such as common [3]. Potamogeton acutifolius, P. densus, P. fi li- It is apparent that many insects are formis, P. frisii, P. gramineus, P. pusillus, capable of dispersing and recolonis- P. rutilus and P. zosterifolius are in a very ing if the environmental conditions vulnerable position. The populations in improve. However, the disappearance Denmark and most other parts of NW of some species could prove irrevers- Europe are few and small (page 109). ible, once the species have disappeared River drop-wort (Oenanthe fl uviatilis) from the entire country or some of the and water-plantain (Elisma natans) are main islands (Funen and Zealand). even more threatened. It is probably Peter Wiberg-Larsen and colleagues unlikely that these species will become have shown that rare species have more widespread and abundant in recolonised new reaches in streams on Danish streams just by natural means Funen [4]. They have also emphasised of dispersal, and we are uncertain as that species that are extinct on Funen to whether this can happen at all. For- are unlikely to arrive from distant tunately, analysis of a recent national populations in Jutland by travelling The large stonefl y Dinocras cephalotes is mapping of the distribution of stream many kilometres over open terrain extinct in Denmark. species will reveal future needs. and crossing the Little Belt to reach the former habitats in the Odense Stream. Will rare insects return? The journey is simply too long and In Denmark, several species of cad- dangerous (page 80). disfl y, damselfl y, mayfl y and stonefl y have become rare over the past 100 years and some species are now extinct (Table 13.1). Among the 42 species of mayfl y recorded from Denmark, fi ve species have become extinct, and fi fteen species are currently threatened 142 RUNNING WATERS

in morphology and physiology, they stream temperature, water chemistry exhibit very close genetic relationships and food supply. as detected by molecular markers, and The recent increase in the size of appear to have evolved within a very salmon stocks in the River Skjern short evolutionary timescale. Nonethe- has been positive. The catch reached less, the North Sea houting is still to 1,000 individuals in 2000, which is be considered unique because of the the largest numbers since the 1940s. distinctive number of gill rakers, the Prospects for the future are promising, elongated upper jaw, the anadromous particularly because of the ongoing life cycle and the ability to withstand large-scale restoration of the River high salinities. The North Sea hout- Skjern (page 30). Magnifi cent progress North Sea houting is an endemic species in ing has been hatched artifi cially and in salmon population recovery would the rivers of SW Jutland. introduced in substantial numbers in be possible if major parts of the river the Vidå Stream and adjacent rivers system, which are not accessible at entering the Wadden Sea. The popula- the moment due to trout farms, were tions have increased, and the species is opened. Freshwater fi sh farms could be Future development of North Sea no longer endangered. abandoned or moved from headwaters houting and salmon Salmon used to be common in to downstream sites or coastal waters. The Danish fauna includes 38 species several rivers in West Jutland (e.g. Although salmon is still present in the of native freshwater fi sh and nine intro- Ribe, Skjern, Sneum, Varde og Vid) Danish fauna, the species is neverthe- duced or escaped species [5]. Among and in the River Gudenå in eastern less rare and the genetic diversity low, the stream species, Denmark has an Jutland. DNA-analysis of dried, old as many ancient, local stocks have international obligation to protect salmon scales has shown that popula- become extinct during the past 100–200 the habitats of rare species, such as tions differed genetically among the years. The native stocks of small brook and river lamprey, salmon and rivers in SW- Jutland, but they were stream, lake and sea trout (Salmo trutta) North Sea houting. Special conserva- more alike than populations from the have experienced the same restriction tion and rehabilitation initiatives have River Gudenå and foreign populations of local genetic diversity. been undertaken in the case of salmon from Scotland and Sweden [8]. The (Salmo salar, page 100) and North Sea populations in the three rivers, Skjern, Extensive help or left undisturbed houting (Coregonus lavaretus) [6]. Varde and Ribe are probably the only There are clear differences in the way North Sea houting was previ- native surviving salmon populations in plants, insects and fi sh in streams are ously distributed throughout the Denmark. Restoration of native Danish managed. Plants and insects are mostly Wadden Sea area of the North Sea salmon stocks aims at ensuring an left undisturbed and they recolonise from the Netherlands to Southern optimal chemical and physical stream restored streams solely by natural Denmark where it spawned in large environment, unrestricted passage of immigration from neighbouring popu- and medium-sized river systems. migrating fi sh and spawning grounds. lations. In contrast, fi sh populations Due to pollution, river regulation and These initiatives have been combined are manipulated to a very great extent. establishment of impassable weirs, the with the introduction of offspring In most Danish streams, trout fry are populations became extinct, except for reared from wild fi sh in farms. Fishery introduced at intervals of 1–4-years a small population in the Vidå Stream. authorities have become very reluctant and their potential competitors or top Genetic studies have suggested a recent to approve the introduction of foreign predators are managed. For the North common origin of all whitefi sh popu- salmon or offspring from other Danish Sea houting and salmon, costly conser- lations (Coregonus sp.) in Denmark, salmon stocks into rivers where native vation and rehabilitation programmes including the North Sea houting, via stocks have (or may have) survived. have been implemented. The interest the postglacial River Elben [7]. While Such native salmon stocks probably in other rare fi sh species e.g. bullhead the Danish whitefi sh populations differ hold essential adaptations to the local (Cottus poecilopus), however, is aston- 13 Streams and their future inhabitants 143

ishingly limited because they do not provide an attractive catch [5]. We feel that future stream and river management should include more general and comprehensive evaluations including their interactions with the catchment, the surrounding valley and the interdependence between plants, invertebrates and fi sh (page 12 and 82). It is unacceptable that a narrow focus on fi shery interests can become a threat to the well-being of plants and invertebrates – or vice versa. We also propose a closer evaluation of whether many years of severe physical regulation, pollution and inappropriate management of the streams in NW Europe have brought some plant and invertebrate species to such a threatened Brown trout in position that they are unlikely to recover and return to Danish streams is improved and restored habitats, where they used to live in the most important large populations only 50–100 years ago. catch for anglers. There are situations where it may be benefi cial to reintro- duce plants or insects to streams. This is especially the case where species have disappeared from a region and hence are native populations which otherwise unable to naturally recolonise sites where they are known to could be lost to fi erce competition and have previously been. Such management could counteract hybridisation with introduced fi sh. the situation we have at present, where the same group of Anglers often want to continue the robust plant and insect species, with a high dispersal capac- tradition of stocking artifi cially reared ity, tend to dominate all streams, irrespective of the present trout fry, even when the density of fry environmental conditions. surpasses the stream carrying capacity It is important, however, that the reintroduction of rare or by several orders of magnitude, lead- extinct plants and insects is performed under strict offi cial ing to enhanced density-dependent control to avoid circumstances where rare species are intro- mortality. duced arbitrarily to generate local publicity. Most botanists and entomologists are against planned re-introductions of New research themes native species, but accept accidental introductions of alien Several authors of preceding chapters species and large-scale dispersal among stream systems by in this book recommend a holistic view weed cutting companies, anglers, biologists and canoeists. of the stream and the catchment from Trout is regarded as the most important Danish stream the spring source to the river mouth species. Nonetheless, it is surprising just how much this spe- (page 12, 26, 45, 66 and 74). Others cies has dominated the discussion and management policy recom mend a stronger focus on the of streams. In most cases, the population density of trout interactions between the organisms and has been the main (or only) criteria for judging the success the physico-chemical environment as of management policy and stream restoration. Continuous well as the mutual interactions between stocking of trout has a selective impact on the biology of the organisms (page 54 and 82). streams and it is surprising that intensive stocking has con- These contributors recommend a tinued even in streams where a self-reproducing population stronger focus on the environment and has been established. The discovery of small native popula- the biota rather than the biology of par- tions of salmon or trout in several Danish rivers has recently ticular species or types of organisms. led to reluctance among fi shery authorities to introduce for- Even the well-being of a single species eign fi sh. There have also been increased efforts to stimulate such as salmon depends on a holistic 144 RUNNING WATERS

activity, could be reduced by re-establishment of meadows and wetlands along the streams. The riparian zone is an integral part of the scenic beauty of streams in the landscape. On closer inspection, the riparian zone may also be important for the biological dynamics, as numerous organisms cross back and forth between land and water during their lifetime. The close interactions suggest that the physical features and vegetation in the riparian zone may be as important for the variety and quality of stream biota as the physico-chemical conditions Comparative stu- within the streams (page 81). We still dies of streams in need to qualify and quantify the impor- different habitats tance of the riparian zone. and geographical Broad-scale ecological research regions can be used also needs to address the importance for drawing general of terrestrial and stream vegetation as conclusions on a source of organic matter that ends biological regulation up being consumed and degraded by in streams. animals, fungi and bacteria in streams. By analysing the proportions of natural isotopes of carbon (C13) and nitrogen attitude towards environmental exploitation and fi shery in (N15) in animals and plants it is possible the sea as well as water quality, food availability, spawning to distinguish between the sources of areas and impassable weirs in the streams (page 92). organic matter from land and water. Holistic studies have two main advantages. They can be Thus the import, export and biological transformed into real environmental policy and manage- conversion of organic matter through ment and they are also easily absorbed and appreciated by the food webs can be traced. This public and administrative circles outside the “world of the approach is likely to reveal exciting dif- stream ecologist”. ferences between streams in forests and Because streams form the network for transport of water more open landscapes and between and nutrients between the terrestrial and aquatic environ- streams of different climates. For exam- ments, it is important to establish the principles for this ple, the food webs in forest streams transport and exchange. This attitude raises questions such can rely entirely on organic matter as: 1) How does stream morphology depend on soil, topog- supplied from the trees e.g. leaves, raphy, vegetation and hydrology?, 2) What are the risks of twigs. In unshaded streams, internal fl oods or prolonged low water fl ows for streams, and 3) production of microalgae and plants What are the patterns and trends of erosion, retention and is more important for the biological transport of mineral particles, organic matter and nutrients transformations. In tropical streams along the streams? (page 26). In Denmark there is an urgent several fi sh species live on seeds and question concerning how much of the nutrient loading, fruits falling from the trees, while fi sh which leads to extensive eutrophication from agricultural species in temperate regions mainly 13 Streams and their future inhabitants 145

live on invertebrates and other fi sh. structure or overall ecosystem function organisms are prevented from moving Interactions between invertebrates and has been of less interest. Also in the freely along the stream. As previously fi sh in tropical and temperate streams management of streams the main focus emphasised it is important for the may, therefore be markedly different. has been on species composition and animal communities that streams work In recent years we have learned relative abundance of fi sh and inver- as networks so that communities at much more about the importance of tebrates and less on overall ecosystem any one site depend on and infl uence the physical environment for the biol- function. We believe that more empha- the composition at neighbouring sites ogy of stream organisms and the role sis will be placed on broad-scale issues through immigration and emigration. of aquatic plants for the fl ow, sediment of ecosystem function in the future as Enclosed reaches lose the essential properties and shelter of invertebrates the natural landscape becomes more feature of streams and increase the risk and fi sh (page 54). However, there is fragmented and human populations in of developing “exaggerated responses” scope for more studies of the effects most countries continue to rise. because animals are unable to leave if of water fl ow and the physical forces We recommend a comprehensive intolerable conditions develop. Biased experienced by stream organisms, review of the control of food webs conclusions for entire stream systems as more advanced instruments and looking at both quantity and qual- can, therefore, be reached based on methods become available. Moreover, ity of basic food sources (“bottom up manipulation of enclosed reaches. stream biologists have not tradition- control”) and the predators (“top down ally dwelled on bio-physical ques- control”) exploiting the basic food In what way are streams a benefi t to tions and have instead concentrated sources. We recommend a broader the community? on the chemical aspects in biology. As consideration of the structure of food For many years the biological, eco- a consequence, we know very little webs, and the conversion of energy logical and aesthetic quality of natu- about the infl uence of fl ow and oxygen through these, to be used, rather than ral habitats has not been suffi ciently concentrations on the respiration and focusing on a few selected interacting appreciated and valued. There was survival of invertebrates on and within species in prey-predator and competi- little concern about the consequences the stream bed. Likewise, although we tor-competitor associations. The wider for the public when ecosystems lost are aware that fl ow and turbulence extrapolation of such latter studies is their natural functions. However, the must be accounted for when plant limited and important compensatory public, is paying for the loss of ecologi- photosynthesis and growth are evalu- responses of other species will be dif- cal services in the streams and their ter- ated as a function of carbon dioxide fi cult to include within the time frame restrial surroundings by enhanced risks concentrations (page 66), we still need commonly used. Under any circum- of erosion and fl ooding, decimation of to establish the relationships under stances spatio-temporal variations in valuable fi sh stocks and pollution of fi eld conditions. the abundance of plants and animals groundwater and coastal waters [9, 10]. are fundamentally controlled by avail- Catastrophic fl ooding, with grave Biological interactions able resources and boundaries set by economic consequences and loss of While the focus has been on the physical forces. We also recommend life, has taken place all over the world importance of the chemical environ- that testable hypotheses are formulated because of channelisation, clear-felling ment for the organisms and less on the a priori to avoid the unfortunate ten- of forests and reclamation of land for physical environment, the biological dency of making post hoc explanations intensive agriculture and industrial or interactions and the ability of organ- to support a preferred concept. urban development in the river valley. isms to infl uence the environment have There are reasons to be sceptical Loss of wetlands has reduced the received very little attention. about the relevance and interpreta- retention of nutrients during transport Fishery biologists have tradition- tion of food web interactions based on of water from the land to the sea. It ally focused on the distribution and small-scale manipulations in enclosed is possible to estimate the value of abundance of economically important stream reaches because it is diffi - reduced nutrient loss, for example, by fi sh species, while the role of fi sh in the cult to ensure the natural dynamics estimating the costs of a similar nutri- control of prey populations, food web and compensatory behaviour when ent reduction in the sewage plants. It 146 RUNNING WATERS

Streams and stream valleys are essential elements of the scenic beauty of open landscapes.

is also possible to estimate the price Future EU policies for the conserva- references for relatively unregulated required for buying land and thereby tion and, if possible, improvement of and unpolluted natural streams. Such putting an end to cultivation of the natural resources and environmental stream systems could stimulate new river valley. Finally, it is possible to quality will be based on the EU Water research, ecological understanding estimate the costs of increased pol- Framework Directive and the EU Habi- and offer exiting opportunities for lution of groundwater and drinking tats Directive. These directives should recreational activities if the streams are water following intensive cultivation of ensure a more sustainable use of the selected and protected at a suffi ciently fi elds along the streams. Scientists have river valley and the survival of rare large scale. Denmark should be able evaluated these ecological services and threatened species. Overall, effec- to promote a few small, species-rich provided by wetlands and streams/ tive nature management must secure lowland streams of high quality. lakes at $14,800 and $8,500/ha per the availability of natural habitats for year in 1997 prices [11]. The main areas plants, invertebrates and fi sh whilst the Adventure, understanding and of expenditure involve prevention of serious eutrophication of streams, lakes excitement erosion and fl ooding, and the introduc- and coastal waters should be reduced. Streams and adjacent areas offer tion of measures to secure the water In future, the public should be able possibilities of excitement and pleas- supply and counteract eutrophication. to enjoy the landscape and the vari- ure, in addition to their ecological On a national scale, Danish society will ous recreational activities associated and economical value. Most people probably pay a high price for a non- with inland waters and should also be fi nd it stimulating to look at a varied chalant attitude towards groundwater provided with drinking water of high landscape and a meandering stream quality as it becomes increasingly quality. rather than a straight channel lined diffi cult and expensive to offer drink- Because virtually all European by monotonous fi elds. This experi- ing water free of pesticides and with lowland streams have been extensively ence is further enhanced if the river nitrate concentrations below the limits impacted by human activity, it will be valley includes a variety of landscape recommended by the World Health especially important to select entire types, such as bogs, fi elds, meadows Organisation. stream systems to serve as future and wetlands. Some appreciate the 13 Streams and their future inhabitants 147

Number of species Insects 627 sharp contour of streams lined with tall within a very short time may seem Flies and mosquitoes 476 Beetles 60 vegetation of shrubs and trees, and wet a high number, taxonomic experts Caddisfl ies 57 meadows generate special experiences have been able to identify about 1,000 Stonefl ies 18 because of their exceptional diversity animal species in a small, unpolluted Mayfl ies 16 of fl owering herbs. temperate stream (Table 13.2). If micro- Roundworms 125 The rich animal life in streams scopic algae and animals are included, Annelids and leeches 56 offers adventure and enlightenment to the experts could probably add an Flatworms 50 every interested person. By sweeping a extra 400 species to the list. Crustaceans 24 kitchen sieve (or pondnet) over the sur- It is an important point to recognise Water mites 22 face of the stream bed or through the that high biological diversity and the Total 904 vegetation, a myriad of animal species living exhibition of the evolution- appear. In clean streams it is possible ary development can be observed by Table 13.2 Number of macroinvertebrate to collect more than 50 different species everybody and is easily accessible species within the most common groups in a within a short time. This collection within small brooks that fl ow through small German stream [12]. will include representatives from most the forest or across the meadow in the larger invertebrate groups that have vicinity of residential areas. Equipped developed during the course of evolu- with a kitchen sieve mounted on a tion. There will be sponges, fl atworms, stick and a jar, a white tray and a few roundworms, annelids, leeches, snails, inexpensive guide books on freshwater lent opportunities for sea trout fi shing clams and crustaceans. And most insect life, children can safely go in search of in most Danish streams and attractive groups, including beetles, caddisfl ies, adventure [13]. opportunities for salmon fi shing in damselfl ies, true fl ies, water bugs, People who prefer larger animals the River Gudenå and the large rivers mayfl ies, stonefl ies and even species of can target a wriggling fi sh on the line in western Jutland. A potential role moth will be represented. and into the frying pan. The excitement model is the Mørum Stream in south- People who are more interested accompanying the fi rst fi sh caught can ern Sweden. This river is an angler’s in the adaptation of species, will fi nd easily match the thrill of a ride on the paradise for trout and salmon fi shing animal species that are adapted to fast roller-coaster in an amusement park, and provides an important income for currents with robust and strong legs, and this excitement of fi shing persists the local community, who are there- claws and suckers that permit secure in adulthood. Anglers are often the fore actively involved in the preven- attachment to the exposed surface of largest interest group working actively tion of acidifi cation, organic pollution stones. Also present will be species in favour of improvement of the envi- and harmful exploitation of the river with large moveable gills that assist ronment and the fi sh populations in system. respiration in slow currents close to streams. Fishing in Danish streams is Streams and rivers have much to the banks or in the vegetation. The often overlooked although the chance offer children of today and adults of tube-living species in the sediments of a successful catch is better here due tomorrow in terms of adventures, are capable of ventilating the tubes and to the much higher density of trout knowledge and positive attitudes tolerating periods of oxygen short- present than in the more famous rivers towards nature. Streams can also age during resting. Species will have of Iceland, Norway and northern become an important source of income different food sources and preferences Sweden. These latter rivers, however, for local communities. There is always encompassing capture of particles by attract more tourists due to the pos- a stream passing close to your home silk meshes, rasping of microalgae on sibility of catching a salmon. – and it is up to you, us and everybody stone surfaces, consumption of dead Fish stocks in Danish rivers could to work for its ecological qualities. or living leaf fragments, consumption be improved even further if a number of fi ne organic particles on the mud of harmful impacts are reduced or surface and the capture of animal prey. eliminated. By removing the presently Although the collection of 50 species impassable weirs there would be excel- 148 RUNNING WATERS

References

1 Lowland river systems Danish stream channels. Technical Miljøundersøgelser. Faglig rapport – processes, form and function Report, no. 6. National Agency of fra DMU nr. 340. Environmental Protection, Freshwa- 2. Kronvang, B., Ærtebjerg, G., Grant, 1. Falkum, Ø., Kronvang, B. & ter Laboratory, Denmark R., Kristensen, P., Hovmand, M., & Svendsen, L.M. (1997) Stoftilbage- 12. Brown, A.V. & Brussock, P.P. (1991) Kirkegaard, J. (1993) Nationwide holdelse på over svømmede enge. Comparisons of benthic inverte- monitoring of nutrients and their Vand og Jord, 4: 125-129. brates between riffl es and pools. ecological effects. State of the 2. Hoffmann, C.C., Pedersen, M.L., Hydrobiologia, 220: 99-108. Danish aquatic environment. Kronvang, B. & Øvig, L. (1998) 13. Pedersen, M.L. (2003) Physical Ambio 22: 176-187. Restoration of the Rivers Brede, habitat structure in lowland streams 3. Larsen, S. E., Kronvang, B., Ovesen, Cole and Skerne. A Joint Danish and effects of disturbance. PhD N.B., and Christensen, O.B. and British EU-LIFE Demonstration thesis. National Environmental (2005) Trends in stream runoff in project, IV – Implications for Nitrate Research Institute, Silkeborg, Denmark, Vand og Jord, 1: 8-13. and Iron Transformation. – Aquatic Denmark. 4. Fenger, J. & Frich, P. (2002) Dansk Conservation. Marine and Freshwa- 14. Kronvang, B., Hansen, H.O., tilpasning til et ændret klima. ter Ecosystems 8(1): 223-240. Friberg, N., Larsen, S.E., Fjorback, Danmarks Miljøundersøgelser. 3. Leopold, L.B., Wolman, M.G. & C. & Johnsen, R. (2000) Habitat Faglig rapport fra DMU nr. 401. Miller, J.P. (1964) Fluvial processes surveys as a tool to assess the 5. Andersen, H.E., Kronvang, B., in geomorphology. W.H. Freeman & benefi ts of stream rehabilitation I: Larsen, S.E, Hoffmann, C.C., Company, San Francisco, USA. The physical dimension. Verhand- Jensen, T.S. and Rasmussen, E.K. 4. Allan, J.D. (1995) Stream ecology. lungen der Internationale Verein i- (2006) Climate-change impacts on Chapman & Hall, London, UK. gung für Theoretische und hydrology and nutrients in a 5. Sand-Jensen, K. & Lindegaard, C. Angewandte Limnologie, 27: 1503- Danish lowland river basin. Science (1998) Økologi i søer og vandløb. 1509. of the Total Enviroment (In Press). Gads Forlag, København, Denmark. 15. Svendsen, L.M., Hansen, H.O. 6. Laubel, A.R., Kronvang, B., Hald, 6. Wiberg-Larsen, P., Brodersen, K.P., (1997) Skjern Å. Sammenfatning af A.B. & Jensen, C. (2003) Hydromor- Birkholm, S., Grøn, P.N. & Skriver, J. den eksisterende viden om de phological and biological factors (2000) Species richness and fysiske, kemiske og biologiske infl uencing sediment and phopsho- assemblages structure of Trichop- forhold i den endre del af Skjern Å- rus loss via bank erosion in small tera in Danish streams. Freshwater systemet. Danmarks Miljøunder- lowland rural streams in Denmark. Biology, 43:633-647. søgelser og Skov- og Naturstyrelsen. Hydrological Processes 17: 3443- 7. Schumm, S.A. (1977) The fl uvial 16. Kronvang, B., Svendsen, L.M., 3463. system. John Wiley and Sons, New Ottosen, O., Nielsen, M.B. & 7. Hasholt, B. (1986) Transport af York, USA. Johannsen, L. (1998) Remeandering partikulært materiale i danske 8. Knighton, D. (1998) Fluvial forms of rivers: short-term implications vandløb. In: Hasholt, B. (ed.) and processes – a new perspective. for sediment and nutrient transport. Partikulært bundet stoftransport i Arnold, London, UK. Verhandlungen der Internationale vand og jorderosjon. Nordisk 9. Brush, L.M. (1961) Drainage basin Vereinigung für Theoretische und Hydrologisk Program, NHP- channels and fl ow characteristics of Angewandte Limnologie, 26: 929- Rapport nr. 14, 203-214. selected streams in central Pennsyl- 934. 8. Kronvang, B., Laubel, A. & Grant, vania. U.S. Geological Survey R. (1997) Suspended sediment and Professional Paper, 282-F: 145-181. particulate phosphorus transport 10. Mernild, S.H. (2001) Regime theory 2 Hydrology, sediment trans- and delivery pathways in an arable and river classifi cation – a temporal port and water chemistry catchment, Gelbæk Stream, and a spatial analysis of natural Denmark. Hydrological Processes rivers in Denmark. Unpubl. Master 1. Ovesen, N.B., Iversen, H.L., Larsen, 11: 627-642. thesis, Institute of Geography, S.E., Müller-Wohlfeil, D.-I., Svend- 9. Andersen, H.E. & Svendsen, L.M. University of Copenhagen. sen, L.M., Blicher, A.S. & Jensen, P. (1997) Suspended sediment and 11. Brookes, A. (1984) Recommenda- M. (2000) Afstrømningsforhold i total phosphorus Transport in a tions bearing on the sinuosity of danske vandløb. Danmarks major Danish river. Methods and References 149

estimation of the effects of a coming 3 From spring to river 10. Riis, T. & Sand-Jensen, K. (2001) major restoration. Aquatic Conser- – patterns and mechanisms Historical changes of species vation. Marine and Freshwater composition and richness accompa- Ecosystems 7: 265-276. 1. Vannote, R.L., Minshall, G.W., nying disturbance and eutrophica- 10. Rebsdorf, Aa., Thyssen, N. and Cummins, K.W., Sedell, J.R. & tion of lowland streams over 100 Erlandsen, M. (1991) Regional and Cushing, C.E. (1980) The river years. Freshwater Biology, 46: 269- temporal variation in pH, alkalinity continuum concept. Canadian 280. and carbon dioxide in Danish Journal of Fisheries and Aquatic 11. Sand-Jensen, K. & Rasmussen, L. streams, related to soil type and Sciences, 37: 130-137. (1979) Macrophytes and chemistry land use. Freshwater Biology 25, 2. Allan, J.D. (1995) Stream ecology. in acidic streams from lignite 419-435. Chapman & Hall, London, UK. mining areas. Botanisk Tidsskrift, 11. Kronvang, B., Grant. R., Larsen, 3. Giller, P.S. & Malmqvist, B. 72: 105-112. S.E., Svendsen, L.M. & Kristensen, (1998) The biology of streams and 12. Bijl, L. van der, Sand-Jensen, K. & P. (1995) Non-point-source nutrient rivers. Oxford University Press, Hjermind, A-L. (1989) Photosynthe- losses to the aquatic environment in Oxford, UK. sis and canopy structure of a Denmark. Impact of agriculture. 4. Sand-Jensen, K. & Lindegaard, C. submerged plant, Potamogeton Marine and Freshwater Research 46: (1998) Økologi i søer og vandløb. pectinatus in a Danish lowland 167-177. Gads Forlag, København. stream. Journal of Ecology, 77: 947- 12. Kronvang, B., Jeppesen, E., Conley, 5. Jacobsen, D. & Friberg, N. 962. D.J., Søndergaard, M., Larsen, S.E., (1998) Macro invertebrate commu- 13. Rosenzweig, M.L. (1995) Species Ovesen, N.B. & Carstensen, J. nities in Danish streams: The effect diversity in space and time. (2005) Nutrient pressures and of riparian forest cover. In: K. Sand- Cambridge University Press, UK. ecological responses to nutrient Jensen & O. Pedersen (Eds.) 14. Brönmark, C., Hermann, J., loading reductions in Danish Freshwater biology. Priorities and Malmqvist, B., Otto, C. & Sjöström, streams, lakes and coastal waters. development in Danish research. P. (1984) Animal community Journal of Hydrology 304: 274-288. Gads Forlag, København, Denmark: structure as a function of stream 13. Grant, R., Laubel, A.R., Kronvang, 208-222. size. Hydrobiologia, 112: 73-79. B., Andersen, H.E., Svendsen, L.M. 6. Wiberg-Larsen, P., Brodersen, K.P., 15. Welcomme, R.L. (1979) Fisheries & Fuglsang, A. (1996) Loss of Birkholm, S., Grøn, P.N. & Skriver, J. ecology of fl oodplain rivers. dissolved and particulate phospho- (2000) Species richness and Longman, London, UK. rus from arable catchments by assemblage structure of Trichoptera surface drainage. Water Research in Danish streams. Freshwater 30(11): 2633-2642. Biology, 43: 633-647. 4 Water fl ow at all scales 14. Kronvang, B., Laubel, A., Larsen, 7. Hildrew, A. & Edington, J.M. (1979) S.E. & Friberg, N. (2003) Pesticides Factors facilitating the coexistence 1. Thyssen, N., Erlandsen, M., and heavy metals in Danish of hydropsychid caddis larva Kronvang, B. & Svendsen, L.M. streambed sediment. Hydrobiologia (Trichoptera) in the same river (1990) Vandløbsmodeller – 494: 93-101. system. Journal of Animal Ecology, biologisk struktur og stofom- 15. Christiansen, L.B. and Plesner, T. 48: 557-566. sætning. NPO-forskning, Nr. C 10. (2001) Intersex og andre effekter på 8. Townsend, C. & Hildrew, A. (1976) Miljøstyrelsen. reproduktionssystemet i skalle og Field experiments on the drifting, 2. Denny, M.W. (1988) Biology and bækørred – relationer til østrogener colonization and continuous the mechanisms of the wave-swept og østrogenlignende stoffer. Teknisk redistribution of stream benthos. environment. Princeton University rapport fra Århus Amt. Journal of Animal Ecology, 45: 757- Press, Princeton, USA. 16. Kronvang, B., Svendsen, L.M., 772. 3. Vogel, S. (1994) Life in moving Jensen, J.P. & Dørge, J. (1999) 9. Johansson, M.E. & Nilsson, C. fl uids. Princeton University Press, Scenario analysis of nutrient (1993) Hydrochory, population Princeton, USA. management at the river basin dynamics and distribution of the 4. Allen, J.D. (1995) Stream ecology. scale. Hydrobiologia 410: 207-212. clonal aquatic plant Ranunculus Chapman & Hall, London, UK. lingua. Journal of Ecology 81: 81-91. 150 RUNNING WATERS

5. Smith, J.R. (1975) Tubulence in Revue der gesamten Hydrobiologie, else af vandløb. Miljøprojekter 30. lakes and rivers. Freshwater 84: 497-508. Miljøstyrelsen. Biological Asssociation. Scientifi c 15. Sand-Jensen, K. (1998) Infl uence of 4. Madsen, T.V., Enevoldsen, H.O. & Publication 29, UK. submerged macrophytes on Jørgensen, T.B. (1993) Effects of 6. Carling, P.A. (1992) The nature of sediment composition and near-bed water motion on photosynthesis the fl uid boundary layer and the fl ow in lowland streams. Freshwa- and dark respiration in submerged selections of parameters for benthic ter Biology, 39: 663-679. stream macrophytes. Plant, Cell and ecology. Freshwater Biology, 28: 16. Sand-Jensen, K. & Mebus, J. R. Environment, 16: 317-322. 273-284. (1998) Fine-scale patterns of water 5. Madsen, T.V. & Søndergaard, M., 7. Statzner, B. & Müller, R. (1989) velocity within macrophyte patches (1983) The effects of current Standard hemispheres as indicators in Danish lowland streams. Oikos, velocity on the photosynthesis of of fl ow characteristics in lotic 76: 169-180. Callitriche stagnalis Scop. Aquatic benthic research. Freshwater 17. Sand-Jensen, K. & Pedersen, O. Botany, 15: 187-193. Biology. 21: 445-459. (1999) Velocity gradients and 6. Sand-Jensen, K., Brodersen, P., 8. Morisawa, M. (1968) Streams: their turbulence around macrophyte Madsen, T.V., Jespersen, T.S. & dynamics and morphology. patches in streams. Freshwater Kjøller, B. (1995) Planter og CO2- McGraw Hill, USA. Biology, 42: 315-328. overmætning i vandløb. Vand & 9. Carling, P.A. (1996) In-stream 18. Madsen, T.V., Enevoldsen, H.O. & Jord, 2: 72-77. hydraulics and sediment transport. Jørgensen, T.B. (1993) Effects of 7. Sand-Jensen, K. (1987) Environ- In: G. Petts and P. Calow (Eds.) water velocity on photosynthesis mental control of bicarbonate use River fl ows and channel forms: and dark respiration in submerged among freshwater and marine Selected extracts from the River stream macrophytes. Plant, Cell and macrophytes. In: C.M.M. Crawford Handbook. Blackwell, UK: 160-184. Environment, 16: 317-327. (Ed.) Plant life in aquatic and 10. Wiberg-Larsen, P., Brodersen, K.P., 19. Townsend, C.R., Scarsbrook, M.R. & amphibious habitats. Blackwell Birkholm, S., Grøn, P.N. & Skriver, J. Dolédec, S. (1997) Quantifying Scientifi c Publications, UK: 99–112. (2000) Species richness and disturbance in streams: alternative 8. Madsen, T.V., (1993) Growth and assemblage structure of Trichoptera measures of disturbance in relation photosynthetic acclimation by in Danish streams. Freshwater to macroinvertebrate species traits Ranunculus aquatilis L. in response Biology, 43: 633-647. and species richness. Journal of the to inorganic carbon availability. 11. Iversen, H.L. & Ovesen, N.B. (1997) North American Benthological New Phytologist, 125: 707-715. Vandføringsevne i danske vandløb Society, 16: 531-544. 1976-1995. Danmarks Miljøunder- 20. Willby, N.J., Pygott, J.R. & Eaton, søgelser. Faglig rapport fra DMU nr. J.W. (2001) Inter-relationships 6 The terrestrial life of 218. between standing crop, biodiversity stream-dwelling insects 12. Larsen, T., Frier, J.O. & Vestergaard, and trait attributes of hydrophytic K. (1990) Discharge/stage relations vegetation in artifi cial waterways. 1. Gullefors, B. & Petersson, E. in vegetated Danish streams. In: Freshwater Biology, 46: 883–902. (1993) Sexual dimorphism in W.R. White (Ed.) Proceedings of relation to swarming and pair international conference on river formation patterns in leptocerid fl ood hydraulics. Wallingford, UK: 5 Aquatic plants caddisfl ies (Trichoptera: Leptoceri- 55-62. dae). Journal of Insect Behaviour, 6: 13. Sode, A. (1997) Effekt af strøm- 1. Hansen, K. (Red.) (1981) Dansk 563-577. rendeskæring på vandføringsevne Feltfl ora. Gyldendal, Denmark. 2. Ivanov, V.D. (1997) Vibrations, og sammensætning og mængde af 2. Pedersen, O. & Sand-Jensen, K. pheromones, and communication grøde. Fyns Amt. (1993) Water transport in sub- patterns in Trichoptera. In: 14. Sand-Jensen, K., Andersen, K. merged macrophytes. Aquatic Holzenthal, R. & O.S. Flint (eds.), & Andersen, T. (1999) Dynamic Botany, 44: 385-406. Proceedings of the 8th International properties of recruitment, expan- 3. Kern-Hansen, U., Holm, T.F., Symposium on Trichoptera, Ohio sion and mortality of macrophyte Madsen, B.L., Thyssen, N. & Biological Survey: 183-188. patches in streams. Internationale Mikkelsen, J. (1980) Vedligehold- References 151

3. Madsen, B.L. (1977) Undersøgelser 12. Malmqvist, B. (2000) How does New Zealand streams. Hydrobio- over populationsbevægelser hos wing length relate to distribution logia, 361: 53-65. vandløbsinsekter. Zoologisk Revy, patterns of stonefl ies (Plecoptera) 21. Svensson, B.W. (1974) Population 39: 22-29. and mayfl ies (Ephemeroptera)? movements of adult Trichoptera at a 4. Timm, T. (1994) Reasons for the Biological Conservation, 93: 271- south Swedish stream. Oikos, 25: shift in dominance between 276. 157-175. Simulium (N.) vernum and Simulium 13. Petersen, I., Winterbottom, J.H., (S.) ornatum (Diptera: Simuliidae) Orton, S., Friberg, N., Hildrew, A.G., along the continuum of an unpol- Spiers, D.C. & Gurney, W.S.C. 7 Macroinvertebrates and biotic luted lowland stream. Archiv für (1999) Emergence and lateral interactions Hydrobiologie, 131: 199-210. dispersal of adult Plecoptera and 5. Jackson, J.K. & Resh, V.H. Trichoptera from Broadstone Stream. 1. Friberg, N. & Winterbourn, M.J. (1989) Distribution and abundance Freshwater Biology, 42: 401-416. (1996) Interactions between of adult aquatic insects in the forest 14. Sode, A. & Wiberg-Larsen, P. riparian leaves and algal/microbial adjacent to a northern California (1993) Dispersal of adult Trichop- activity in streams. Hydrobiologia, stream. Environmental Entomology, tera at a Danish forest stream. 341: 51-56. 18: 278-283. Freshwater Biology, 30: 439-446. 2. Kjeldsen, K. (1996) Regulation of 6. Sweeny, B.W. (1993) Effects of 15. Larsen, M.R. (1999) Dispersal of algal biomass in a small lowland streamside vegetation on macroin- adult Trichoptera and Ephemerop- stream: fi eld experiments on the vertebrate communities of White tera in three different types of role of invertebrate grazing, Clay Creek in eastern North vegetation. MS thesis, Freshwater phosphorus and irradiance. America. Proceedings of The Biological Laboratory, University of Freshwater Biology, 36: 535-546. Academy of Natural Sciences of Copenhagen (in Danish with 3. Friberg, N. (1997) Benthic inverte- Philadelphia, 144: 291-340. English abstract). brate communities in six Danish 7. Müller, K. (1982) The colonization 16. Kovats, Z.E., Ciborowski, J.H. & forest streams: Impact of forest type cycle of Freshwater Insects. Corkum, L.D. (1996) Inland on structure and function. Ecogra- Oecologia, 52: 202-207. dispersal of adult aquatic insects. phy, 20: 19-28. 8. Madsen, B.L. & Butz, I. (1976) Pop- Freshwater Bio logy, 36: 265-276. 4. Feminella, D.M., Power, M.E. & ulation movements of adult 17. Griffi th, M.B., Barrows, E.M. & Resh, V.H. (1989) Periphyton Brachyptera risi (Plecoptera). Oikos, Perry, S.A. (1998) Lateral dispersal responses to invertebrate grazing 27: 273-280. of adult aquatic insects (Plecoptera, and riparian canopy in three 9. Madsen, B.L., Bengtsson, J. & Butz, Trichoptera) following emergence northern California coastal streams. I. (1977) Upstream movement by from headwater streams in forested Freshwater Biology, 22: 445-457. some Ephemeroptera species. Archiv Appalachian catchments. Annals of 5. Jacobsen, D. & Sand-Jensen, K. für Hydrobiologie, 81: 119-127. the Entomological Society of (1994) Invertebrate herbivory on the 10. Petersen, I., Winterbottom, J.H., America, 91: 195-201. submerged macrophyte Potamoge- Orton, S. & Hildrew, A.G. 18. Malmqvist, B., Rundle, S., Brön- ton perfoliatus in a Danish stream. (1999) Does the colonization cycle mark, C. & Erlandsson, A. Freshwater Biology, 31: 43-52. exist? In: N. Friberg & J.D. Carl (1991) Invertebrate colonization of 6. Iversen, T.M., Thorup, J., Hansen, (eds.) Biodiversity in Benthic a new, man-made stream in T., Lodal, J. & Olsen, J. (1985) Quan- Ecology. Proceedings from Nordic southern Sweden. Freshwater titative estimates and community Benhtological Meeting in Silkeborg, Biology, 26: 307-324. structure in a macrophyte rich Denmark, 13-14 November 1997. 19. Wiberg-Larsen, P., Bennike, O., stream. Archiv für Hydrobiologie, National Environmental Research Jensen, J.B. & Lemke, W. (2001) Tri- 102: 291-301. Institute, Denamrk. NERI Technical choptera remains from early 7. Andersen, T.H., Friberg, N., Report No. 266: 123-126. Holocene river deposits in the Great Hansen, H.O., Iversen, T.M., 11. Anholt, B.R. (1995) Density Belt, Denmark. Boreas, 30: 299-306. Jacobsen, D. & Krøjgaard, L. dependence resolves the stream 20. Collier, K.J. & Smith, B.J. (1998) Dis- (1993) The effects of introduction of drift paradox. Ecology, 76: 2235- persal of adult caddisfl ies (Trichop- brown trout (Salmo trutta L.) on 2239. tera) into forest alongside three Gammarus pulex L. drift and density 152 RUNNING WATERS

in two fi shless Danish streams. 7. Dieperink, C. & Rasmussen, P.C. 2. Baagøe, J. (1897) Potamogeton Archiv für Hydrobiologie, 126: 361- (1996) Reduction of brown trout undulatus Wolfgang (P. crispus L. X 371. (Salmo trutta L.), salmon (S. salar L.) praelongus Wulf.). Botanisk 8. Friberg, N., Andersen, T.H., Hansen, and rainbow trout (Oncorhynchus Tidsskrift, 21. H.O., Iversen, T.M., Jacobsen, D., mykiss (Walbaum)) smolt bycatches 3. Raunkiær, C. (1895-99) De danske Krøjgaard, L. & Larsen, S.E. in eel pound nets. Fisheries blomsterplanters naturhistorie. (1994) The effect of brown trout Management and Ecology, 3: 101- Enkimbladede. Gyldendal, Køben- (Salmo trutta L.) on stream inverte- 109. havn, Denmark. brate drift, with special reference to 8. Rasmussen, G. (1995) Hvad bliver 4. Rosenzweig, M.L. (1995) Species Gammarus pulex L. Hydrobiologia, der af smoltene? Årsberetning 1994 diversity in space and time. 294: 105-110. fra Institut for Ferskvandsfi skeri og Cambridge University Press, 9. Hildrew, A.G. (1996) Food webs Fiskepleje, IFF: 23-34. Cambridge, UK. and species interactions. In: G. Petts 9. Jepsen, N., Aarestrup, K. & Rasmus- 5. Henry, C.P., Amoros, C. and & P. Calow (Eds.) River Biota, sen, G. (1997) Smoltdødeligheder i Bornette, G. (1996) Species traits diversity and dynamics. Blackwell Tange Sø undersøgt i foråret 1996. and recolonization processes after Science, Oxford, UK: 123-144. Danmarks Fiskeriundersøgelser fl ood disturbances in riverine rapport nr. 32. macrophytes. Vegetatio, 122, 13-27. 10. Larsen, K. (1959) The effect of the 6. Windolf, J., Svendsen, L.M., Skriver, 8 Stream fi sh and desirable liberation of sea-trout fry in the J., Ovesen, N.B., Larsen, S.E., fi sh stocks Gudenå area, as shown by the trout Iversen, H.L., Erfurt, J. and Müller- catch in the lower river and the Wohlfeil, D. (1998) Ferske Vandom- 1. Johansen, A.C. & Løfting, J.C. Randers Fjord. Rapports et Proces- råder – Vandløb og kilder. (1919) Fiskebestanden og fi skeriet i Verbaux des Reunions. Conceil Vandmiljøplanens Overvåg nings- Gudenåens nedre løb og Randers International pour l’Exploration de program 1997. Danmarks Fjord I. Skrifter udgivne af Kommis- la Mer, 148: 26-28. Miljøundersøgelser. Faglig rapport sionen for Havundersøgelser no. 9. 11. Nielsen, E.E., Hansen, M.M. & fra DMU nr. 253. Bianco Lunos Bogtrykkeri, Køben- Loeschke, V. (1997) Analysis of 7. Grime, J.P. (1979) Plant Strategies havn, Denmark. microsatellite DNA from old scale and Vegetation Processes. John 2. Poulsen, E.M. (1935) Nye under- samples of Atlantic salmon Salmo Wiley & Sons, Chichester, UK. søgelser over Gudenåens Lakse- og salar: A comparison of genetic 8. Newbold, C. and Palmer, M. (1979) havørredbestand. Beretning til composition over 60 years. Molecu- Trophic adaptations of Aquatic Ministeriet for Landbrug og Fiskeri. lar Ecology, 6: 487-492. Plants. CST Note No. 18, Nature Dansk Biologisk Station, XL: 9-36. 12. Koed, A., Aarestrup, K., Nielsen, Conservancy Council, England. 3. Otterstrøm, C.V. (1936) Turbinerne E.E. & Glüsing, H. (1999) Status for 9. Riis, T. and Sand-Jensen, K. (2001) og de nedvandrende ungfi sk af laks laksehandlingsplanen. Danmarks Historical changes in species og ørred (samt ål) III. Sportsfi sk- Fiskeriundersøgelser rapport nr 66. composition and richness accompa- eren, 12: 131-142. 13. Hickley, P. & Tompkins, H. (Eds.) nying pertubation and eutrophica- 4. Nielsen, J. (1985) Havørreden i (1998) Recreational Fisheries, social, tion of Danish lowland streams over Gudenåen. Gudenåkomiteen. economic and management aspects. 100 years. Freshwater Biology, 46, Rapport nr. 3. Århus, Viborg og Fishing News Books, Blackwell 269-280. Vejle Amtskommuner. Science, Oxford, UK. 10. Baattrup-Pedersen, A., Riis, T. and 5. Nielsen, J. (1986) Laksefi skene og Hansen, H.O. (1998) Grødeskæring fi skeriet i Randers Fjord. Gudenå- og vandplanter i danske vandløb. komiteen. Rapport nr. 4. Århus, 9 Water plants past and present Vand & Jord, 4, 136-139. Viborg og Vejle Amtskommuner. 11. Sand-Jensen, K., Riis, T., Vester- 6. Rasmussen, P.C. (1992) Fiskeri og 1. Baagøe, J. and Ravn, F.K. (1896) gaard, O. and Larsen S.E. (2000) laksefi sk – Randers Fjord 1990-91. Excursioner til jydske søer og Macrophyte decline in Danish lakes IFF rapport nr 6. Danmarks vandløb i sommeren 1895. Botanisk and streams over the past 100 years. Fiskeriundersøgelser. Tidsskrift, 20, 288-326. Journal of Ecology, 88, 1030-1040. References 153

10 Improvements ahead for Denmark. Water Research 18 (2): 20. Wiberg-Larsen, P., Petersen, S.E., macroinvertebrates ? 145-151. Madsen, H.B., Larsen, F.G., Knud- 13. Kirkegaard, J., Wiberg-Larsen, P., sen, J. & Adamsen, N.B. 1. Wesenberg-Lund, C. (1940) Det Jensen, J., Iversen, T.M. & (1994) Renere fynske vandløb. Ferskvandsbiologiske Laboratorium Mortensen, E. (1992) Biologisk Vand & Jord, 1: 10-13. gennem 40 år. Bianco Lunos bedømmelse af vandløbskvalitet. 21. Wiberg-Larsen, P., Larsen, F.G., Bogtrykkeri, København, Denmark. Metode til anvendelse på stationer i Knudsen, J. & Adamsen, N.B. 2. Stoltze, M. & Pihl, S. (Eds.) (1998) Vandmiljøplanens Overvågnings- (1994) Rent vand – ikke bare en Rødliste 1997 over planter og dyr i program. Danmarks Miljøunder- døgnfl ue? Vand & Jord, 1: 62-64). Danmark. Miljø- og Energiministe- søgelser. Teknisk anvisning fra 22. Aagaard, P. & Bolet, B. (1997) riet, Danmarks Miljøundersøgelser DMU nr. 5. Danmarks største slørvinge igen og Skov- og Naturstyrelsen (in 14. Skriver, J., Friberg, N. & Kirkegaard, udbredt i Storå. Vand & Jord, 4: 32- Danish with English summary). J. (2001) Biological assessment of 35. 3. Jensen, C.F. (1984) De danske running waters in Denmark: Baetis-arter (Ephemeroptera: introduction of the Danish Stream Baetidae). Flora & Fauna, 90: 97-102 Fauna Index (DSFI). Verhandlungen 11 A new development: (in Danish with English summary). der Internationale Vereinigung für Stream restoration 4. Jensen, F. Personal communication. Theoretische und Angewandte 5. Findal, J.K. (1916) Århus Å. Bidrag Limnologie 27: 1822-1830. 1. Viborg Amt (1993) Miljøtilstanden i til en undersøgelse af Områdets 15. Skriver, J. (2001) Biologisk vandløbs- vandløbene i Spøttrup kommune. Fauna. Flora & Fauna, 1916: 1. kvalitet (DVFI). In: Bøgestrand, J. Rapport nr. 116 i miljøserien. Hefte. (ed.): Vandløb og kilder 2000. 2. Frandsen, S.B., Levesen, B., Jes- 6. Munk, T. (1984) Døgnfl uen NOVA 2003. Danmarks persen, J.W. & Rasmussen, K. Rhitrogena germanica Eaton i Højen Miljøundersøgelser. Faglig rapport (1993) Renere vandløb i Vejle Bæk. Flora & Fauna, 90: 103-105. fra DMU nr. 378: 61-64. Amt. Vand & Miljø, 10: 123-125. 7. Jensen, C.F. (1951) Plecoptera 16 Iversen, T.M., Madsen, B.L. & 3. Smedegaard-Mortensen, J. (1963) (Slørvinger). En faunistisk biologisk Bøgestrand, J. (2000) River conser- Østenstormen svejsede sammen. undersøgelse af Skjern Å. I. Flora & vation in the European Community, Hedeselskabets Tidsskrift, 1: 27-32. Fauna, 1951: 17-40 (in Danish with including Scandinavia. In: Boon, P. 4. Rasmussen, J.B. (1999) Skjern Å English summary). J., Davies, B.R. & Petts, G.E. (Eds.). Naturprojektet. Miljø- og 8. Wiberg-Larsen, P. (1986) Holocentro- Global perspectives on River Energiministeriet, Skov- og pus insignis Martynow, 1924 and Conservation: Science, Policy and Naturstyrelsen. Micrasema setiferum (Pictet, 1834) Practice. John Wiley & Sons Ltd.: 79- 5. Skov- og Naturstyrelsens Skjern Å (Trichoptera) new to Denmark. 103. hjemmeside: www.sns.dk/natur/ Entomologiske Meddelelser 53: 53- 17. Miljøstyrelsen (1997) Punktkilder. netpub/skjernaa/index.htm 57. Orientering fra Miljøstyrelsen nr. 6. COWI Consult & Skov- og 9. Wiberg-Larsen, P. (1985) Revision 16/1997. Miljø- og Energiminis- Naturstyrelsen (1999) Kort over of the Danish Hydroptilidae teriet. Skjern Å Naturprojektet, Fremtidig (Trichoptera). Entomologiske 18. Baattrup-Pedersen, A. Skriver, J. & vegetation. Meddelelser 53: 39-45. Wiberg-Larsen, P. (2001) Weed- 7. Hansen, H.O. (Red.) (1996) 10. Wiberg-Larsen, P. Personal commu- cutting practice and impact on trout Vandløbsrestaurering – eksempler nication. density in Danish lowland streams. og erfaringer fra Danmark. Dan- 11. Landbrugsministeriet (1970) Verhandlungen der Internationale marks Miljøunder søgelser. Faglig Vejledning om fremgangsmåden Vereinigung für Theoretische und rapport fra DMU nr. 151. ved bedømmelse af recipienters Angewandte Limnologie 27: 674- 8. Sand-Jensen, K. (1997) Macro- forureningsgrad. 677. phytes as biological engineers in the 12. Andersen, M.M., Riget, F.F. & 19. Wiberg-Larsen, P., Petersen, S., ecology of Danish streams. In: K. Sparholt H. (1984) A modifi cation Rugaard, T. & Geertz-Hansen, P. Sand-Jensen & O. Pedersen (Eds.) of the Trent Index for use in (1994) Bedre vandløbspleje giver Freshwater biology – Priorities and fl ere fi sk. Vand & Jord, 1: 263-265. development in Danish research. G. 154 RUNNING WATERS

E.C. Gad, København, Denmark: 9. Kaiser, E.W. (Red.) (1949) Fysisk, 13 Streams and their future 74-101. kemisk, hydrometrisk og biologisk inhabitants 9. Sode, A. (1997) Effekt af undersøgelse af Mølleåen fra strømrendeskæring på vandføring- Lyngby Sø til Øresund 1946 og 1. Mackay, R.J. (1992) Colonization by sevne og sammensætning og 1947. Dansk Ingeniørforening, lotic macroinvertebrates: A review mængde af grøde i Ringe Å. Fyns Spildevandskomiteen. Skrift, 1: 1-39. of processes and patterns. Canadian Amt, Natur- og Vandmiljøafdelin- 10. Kaiser, E.W. (Red.) (1951) Biolo- Journal of Fisheries and Aquatic gen. giske, biokemiske, bakteriologiske Sciences, 49: 617-628. 10. Baattrup-Pedersen, A., Riis, T. & samt hydrometriske undersøgelser 2. Stolze, M. & Pihl, S. (1998) Rødliste Hansen, H.O. (1998) Grødeskæring af Pøleåen 1946 og 1947. Dansk 1997 over planter og dyr i Danmark. og vandplanter i danske vandløb. Ingeniørforening, Spildevands- Miljø- og Energiministeriet, Vand & Jord, 5: 136-139. komiteen. Skrift, 3: 1-53. Danmarks Miljøundersøgelser og 11. Fjerdingstad, E. (1950) The micro- Skov- og Naturstyrelsen. fl ora of the River Mølleaa. With 3. Skriver, J., Jensen, F., Bundgaard, P. 12 Environmental state Special reference to the relation of & Holm, P. (2005) The stonefl y and research the benthic algae to pollution. Folia Perlodes microcephalus expanding its Limnologica Scandinavica, 5: 1-123. range in Denmark. Flora & Fauna, 1. Miljøstyrelsen (1998) Biologisk 12. Moss, D., Furse, M.T., Wright, J.F. & 111: 95-103 (in Danish with English bedømmelse af vandløbskvalitet. Armitage, P.D. (1987) The predic- summary). Vejledning fra Miljøstyrelsen nr. 5. tion of the macro-invertebrate fauna 4. Wiberg-Larsen, P., Pedersen, S.E., 2. Skriver, J. (1999) Biologisk vandløbs- of unpolluted running-water sites Tornberg, N.H., Sode, A., Muus, K. kvalitet. In: J. Bøgestrand (Red.) in Great Britain using environmen- & Wehrs, M. (1997) De fynske Vandløb og kilder 1998. Danmarks tal data. Freshwater Biology, 17: 41- vandløb. Vandmiljøovervågning. Miljøundersøgelser. Faglig rapport 52. Tema. Ferskvand. Fyns Amt, Natur- fra DMU nr. 292: 53-57. 13. Wright, J.F., Armitage, P.D., Furse, og Vandmiljøafdelingen. 3. Skriver, J., Baattrup-Pedersen, A. & M.T. & Moss, D. (1988) A new 5. Ernst, M.E. & Nielsen, J. (1981) Larsen, S.E. (1997) Vandløbenes approach to the biological surveil- Sjældne og truede ferskvandsfi sk i miljøtilstand. I: J. Windolf (Red.) lance of river quality using mac- Danmark. Meddelelser fra Fersk- Ferske vandområder. Vandløb og roinvertebrates. Verhandlungen der vandsfi skerilaboratoriet 1/81. kilder. Vandmiljøplanens Over- Internationale Vereinigung für 6. Jensen, A.R., Nielsen, H.T. & Ejbye- vågningsprogram 1996. Danmarks Theoretische und Angewandte Ernst, M. 2003 National Manage- Miljøundersøgelser. Faglig rapport Limnologie, 23: 1548-1552. ment Plan for the Houting. Ministry fra DMU nr. 214: 29-46. 14. Miljøstyrelsen (1998) Punktkilder of the Environment, Forest and 4. Stoltze, M. & Pihl, S. (1998) Rødliste 1997. Orientering fra Miljøstyrelsen Nature Agency, The County of 1997 over planter og dyr i Danmark. nr. 9. Sønderjylland and the County of Miljø- og Energiministeriet. 15. Olesen, N.K. & Roth, E. (1999): Det Ribe. http://www2.skovognatur. Danmarks Miljøundersøgelser og danske dambrugserhverv. En dk/udgivelser/2004/Forvaltnings- Skov- og Naturstyrelsen. strukturanalyse. IME Report 1/99. plan_for_snaebel_engelsk.pdf 5. Nielsen, J. (1997) Ørreden som Institut for Miljø- og Erhvervs- 7. Hansen, M.M., Mensberg, K.-L. D. miljøindikator. Miljønyt nr. 24. økonomi, Syddansk Universitet, & Berg, S. (1999) Postglacial Miljø- og Energiministeriet. Esbjerg, Denmark. recolonization patterns and genetic 6. Berg, K. (1943) Physiographical 16. Iversen, T.M. (1995) Fish Farming relationships among whitefi sh Studies on the River Susaa. Folia in Denmark. Environmental impact (Coregonus sp.) populations in Limnologica Scandinavica, 1: 1-74. of regulative legislation. Water Denmark, inferred from mitochon- 7. Berg, K. (Ed.) (1948) Biological Science and Technology, 31: 73-84. drial DNA and microsatellite Studies on the River Susaa. Folia markers. Molecular Ecology 8, 239- Limnologica Scandinavica, 4: 1-138. 252. 8. Wesenberg-Lund, C. (1925) Om 8. Nielsen, E.E., Hansen, M.M. & forureningen af vore ferske vande. Loeschke, V. (1997) Analysis of Ferskvandsfi skeribladet, 23: 157-163. microsatellite DNA from old scale References 155

samples of Atlantic salmon Salmo salar: A comparison of genetic composition over 60 years. Molecu- lar Ecology, 6: 487-492. 9. Friberg, N. & Carl, J.D. (Eds.) (1997) Biodiversity in Benthic Ecology. Proceedings from Nordic Bentho- logical Meeting in Silkeborg, Denmark, 13-14 November 1997. National Environmental Research Institute, Denmark. NERI Technical Report No. 266. 10. Sand-Jensen, K. (2000) Fresh waters: Human impact on. Encyclo- pedia of Biodiversity. Academic Press. 11. Costanza, R., d’Arge, R., de Groot, R., Farber, S., Grasso, M., Hannon, B., Limburg, K., Naeem, S., O’Neil, R.V., Paruelo, J., Raskin, R.G., Sutton, P. & van den Belt, M. (1997) The value of the world’s ecosystem services and natural capital. Nature, 387: 253-260. 12. Allan, J.D. (1995) Stream Ecology. Chapman & Hall, London, UK. 13. Olsen, L.H. (1999) Små dyr i sø og å. Gads Forlag, København, Denmark. 156 RUNNING WATERS

Index

A biodiversity 25, 102, 127, 129, 133 Cottus biological interactions 83, 137, 145 C. poecilopus 142 abstraction of water 139 biota 24, 83, 143, 144 county 118, 120, 126, 127, 134, 137 accumulation 35, 58 biotic 17, 21, 83, 90, 91, 118 current velocity 17-22, 59, 65, 101, 102, acidifi cation 29, 35, 43, 140, 147 birds 78, 97, 131, 140 136 acid rain 35 blackfl y 78 adaptations 63-65, 75, 142 BOD 28, 39 D administration 99 Bolbro Brook 33 Agapetus dam 95, 101 Bornholm 14, 30, 33, 53, 99 A. fuscipes 77, 79 damselfl y 74, 141 bottom substratum 18, 46 A. ochripes 80 Danish Stream Fauna Index 117, 118, boundary layer 70-72 agriculture, outlets from 16, 22, 26, 35, 120, 134 Brachycentrus 36, 38, 40, 41, 123, 124, 128, 130, 133, deforestation 10 B. subnubilus 80 139, 140, 145 detritivorous 84 Brachyptera alder 81, 85 Dinocras B. braueri 116 algae 46, 47, 68, 77, 81, 84, 85, 91, 147 D. cephalotes 116, 141 B. risi 77 Alisma dipper 67 Brede Stream 14, 15, 34, 122, 126-129 A. gramineum 112 Diptera 75 broodstock 98, 103 aluminium 35 discharge 13, 16-20, 27, 32, 40, 47, 53, brook 14, 16, 79, 87, 97, 142 Anabolia 55-60, 85, 88, 91, 98, 101, 117, 119, bryophyte 72 A. furcata 80 124, 130, 136, 140 A. nervosa 80, 81, 86 dispersal 49, 50, 52, 60, 77-81, 106-113, C Ancylus 141, 143 A. fl uviatilis 85 caddisfl y 79, 80, 86, 87, 91, 141 dispersal index 108, 111 anoxic 29, 73, 107 Callitriche 51, 52, 60-65, 71, 107, 111, distribution 18, 28, 48-52, 55, 57, 59, 61, Ansager Stream 34 112 69, 73, 75, 80, 108, 113, 116, 121, 134, arable land 30, 36 C. cophocarpa 51, 65, 111 137, 139, 141, 145 assessment of water quality 70, 71, C. platycarpa 51, 61, 111 disturbance 23, 49, 60, 63-65, 85, 106- 117, 118, 120, 134 Calopteryx 115, 140 C. virgo 74, 120 diversity 15, 23-25, 47, 57, 65, 87, 102, B Capnia 106, 107, 113, 116, 118, 124, 136, 142, bacteria 84, 144 C. bifrons 82 147 Baetis carbon dioxide 58, 70-73, 145 DNA-analyses 99 B. buceratus 116 catchment area 15, 17, 32, 34, 73 Dolania B. digitatus 116 Ceratophyllum D. americana 78 B. muticus 116 C. demersum 49, 112 domestic 41, 103, 139 Bagge Stream 33 channelisation 19, 23, 29, 35, 55, 124, drag 56, 59-62, 64, 140 Bangsbo Stream 116 145 dragonfl y 120 beetle 78 channel regulation 106 drain 37, 55, 58, 94, 106, 123, 126, 128 benthos 87, 89 Chironomidae 77 drainage 14, 23, 29, 31, 33-37, 41, 93, Beraea chlorophyll 85, 137 101-105, 115, 130, 131, 136, 139 B. pullata 77 colonisation 68-73, 78-80, 113, 140 drift 78, 87-89 Berula competition 83, 90, 91, 99, 143 drought 30 B. erecta 68, 71, 112 Copenhagen 31, 105, 135, 139 bicarbonate 52, 71-73 Coregonus C. lavaretus 142 Index 157

E G Isoptena I. serricornis 141 ecological status 121 Gammarus Ithytrichia ecology 83, 86, 113, 135 G. pulex 87, 89 I. lamellaris 77 ecosystem 16, 25, 45, 84, 85, 90, 91, 133, Gelbæk Stream 34 136, 145 Gelså Stream 23, 24, 86, 87 J eel 94-101 Germany 94, 103, 118, 123 egg 77- 79, 102 Gjern Stream 15, 116 Jeksen Brook 116 electrofi shing 98 glaciation 10, 13, 35, 53, 80, 93 Jutland 13-16, 19, 30-38, 40, 41, 43, 53, Elisma glacier 13 56, 61, 64, 80, 85, 86, 93, 94, 100, 116, E. natans 141 glass eel 94, 100, 101 120, 121, 126, 130, 134, 136, 141, 142, Ellebæk Brook 33 grazers 85, 91 147 Elodea Grejs Stream 116 K E. canadensis 51, 60, 71, 111, 112 Grenå Stream 17 emergent plants 59 Grindsted Stream 34 Karup Stream 17, 96, 99, 141 Ephemera 114, 141 groundwater 16, 19, 27, 29-31, 34-38, Kolding Stream 116 erosion 13, 18, 19, 23, 33, 34, 39, 42, 46, 40, 41, 47, 48, 101, 126, 130, 131, 139, Konge Stream 17 49, 57, 60, 101, 106, 119, 140, 144, 145, 146 L 145, 146 Grønå Stream 34 estuary 95-100 guild 85 laminar fl ow 55-58, 65 Eusimulium gut 89 lamprey 142 E. vernum 78 Lasiocephala H eutrophication 40, 49, 106-111, 144, 146 L. basalis 116 eutrophication index 108 habitat variation 20, 107 leachate 84 eutrophic species 107, 108 Halleby Stream 17 leaching 29, 38, 39, 85 Hansted Stream 116 legislation 103 F hatchery 99, 103 Lellinge Stream 116 fall 85 heavy metals 34, 35, 39-41, 115 Leuctra feeding 75, 77, 81, 87, 116 Heptagenia 80, 116, 121, 141 L. fusca 80, 121 fertiliser 28, 36, 107, 130 H. sulphurea 80, 121, 141 life cycle 75, 99, 142 fi ne particulate organic matter 15, 46 heterogeneity 21, 46-48 Lillebæk Brook 33 fi sh farms 27, 117, 120, 124, 136, 137, hormone 41, 101 Lilleå Stream 116 139, 142 hydroelectric power 124, 127 Limnephilidae 77 fi shing 22, 94-97, 101-103, 136, 147 hydrological cycle 30, 38 Lindenborg Stream 17, 116 fi sh ladder 95-98 hydrology 25, 28, 31, 53, 102, 139, 144 litter 140 fjord 130 Hydropsyche 79 lotic 80, 83 fl ooding 14, 15, 46, 58, 126, 128, 130, Hydroptila 79, 116 Luronium 131, 145, 146 Højen Brook 116 L. natans 112 fl oodplain 14, 15, 23-25, 29, 41, 46, 126 Højvads Rende Stream 33 Lyby-Grønning Stream 33 food web 78, 90, 91, 144, 145 Lyngbygaard Stream 34 I fry 95-100, 142, 143 Lype Funder Stream 115, 116 intermediate disturbance theory 65 L. reducta 79 Funen 31, 33, 53, 78-81, 121, 134, 137, intersex 41 141 iron 29, 35, 36, 87 fungi 77, 144 158 RUNNING WATERS

M Odontocerum P. fi liformis 110, 111, 141 O. albicorne 116 P. frisii 51, 141 macroinvertebrate 20, 75, 84-91, 115- Oenanthe P. gramineus 111, 141 121, 134, 136, 147 O. fl uviatilis 112, 141 P. lucens 51, 110 macrophyte 49, 52, 86, 111, 119, 135 Omme Stream 131 P. natans 51, 110 Maglemose Stream 33 organic matter 15, 27, 29, 37, 45, 46, P. pectinatus 51, 111 malaise trap 78, 79, 81 67, 70, 78, 84, 86, 91, 117, 119, 130, P. perfoliatus 51, 110 management 43, 60, 90, 98-106, 119, 136, 144 P. polygonifolius 111 133, 135, 137, 139, 140, 143-146 Orten Brook 116 P. praelongus 51 Manning number 56 Osmylus P. pusillus 141 Mattrup Stream 34, 86, 116 O. fulvicephalus 117 P. rutilus 111, 141 mayfl y 78, 114, 121, 141 otter 131 P. zosterifolius 110, 111, 141 meadow 130, 147 oxygen 23, 45, 56, 58, 62, 63, 65, 69, 70, Potamophylax meander 15, 19, 127 73, 101, 102, 117, 118, 139, 145, 147 P. cingulatus 79 mean discharge 58 precipitation 29-38, 41, 101 mean velocity 55-59, 65 P predation 87, 89, 91, 97 Micrasema pyrite 29, 35, 36, 42, 119 M. setiferum 116 Paraleptophlebia Micropterna P. submarginata 141 Q M. sequax 77 perch 94, 101 midge 118 Perlodes quality of streams 90, 135, 136, 139, Ministry of the Environment 27, 129 P. dispar 116, 141 140 mink 101 P. microcephalus 118, 121, 141 R morphological index 111 pesticide 41, 137 municipality 25, 97, 119-121, 124, 125, pH 35, 69, 73 rainfall 106, 117 137, 139, 140 pheromones 77 Ranunculus 49, 51, 60, 62, 64, 68, 111, Myriophyllum phosphorus 29, 33-43, 107 112 M. spicatum 49, 112 photosynthesis 46, 48, 61, 65, 70-73, R. aquatile 51 M. verticillatum 49, 112 145 R. baudotii 51 Mørum Stream 147 pike 94, 101 R. circinatus 49 Plecoptera 75, 77, 80 R. fl uitans 51 N Plectrocnemia R. lingua 49 nitrate 38, 41, 42, 107, 146 P. conspersa 79, 87 R. peltatus 51 nitrogen 28, 29, 33, 36-43, 107, 126, 128, pollution 27 recolonisation 48, 65, 80, 106, 140 144 Polycentropus 79 Red Data Book 116, 121 North Sea 100, 142 pondweed 51, 60, 104, 105, 108-111, reference condition 115 nymph 82 140, 141 rehabilitation 41, 43, 99, 100, 128, 131, pool 19-23, 53, 56, 72, 73, 106, 123 142 O Potamogeton 49, 51, 52, 72, 110-112, 140, respiration 46, 48, 63, 65, 70, 145, 147 oak 94 141 restoration 20, 24, 25, 43, 97, 121, 124- ochre 29, 35, 36, 43, 118, 119, 126, 127, P. acutifolius 111, 112, 141 131, 136, 140, 142, 143 130, 134, 139 P. alpinus 110, 111 Reynolds number 55, 56, 61, 65 Odagmia P. berchtoldii 51 Rhithrogena O. ornata 78 P. coloratus 111 R. germanica 116 Odderbæk Brook 33 P. crispus 51, 110 Ribe Stream 17, 100, 116 Odense Stream 17, 80, 141 P. densus 111, 141 riffl e 19-23, 56, 106, 123, 125 Index 159

riparian vegetation 20, 78, 118 stream order 14, 16-18, 29, 30, 47, 50 Z riparian zone 67, 71, 81, 83-85, 90, 144 suspended solids 19 zander 94-97 River Continuum Concept 45, 46 Suså Stream 17, 34, 50, 86, 135 Zannichellia River Gudenå 10, 13, 17, 22, 32, 34, 48- Svinebæk Brook 14 Z. palustris 60 50, 93-98, 104, 115, 116, 124, 142, 147 swarming 76, 77, 81 Zealand 30, 31, 33, 48, 53, 65, 80, 84, River Skjern 10, 13, 14, 17, 20, 23, 24, Sweden 97, 100, 142, 147 116, 141 29, 32, 34, 99, 100, 103, 115, 116, 125, Sæby Stream 116 126, 130, 131, 141, 142 Ø River Storå 10, 17, 116, 121, 127 T Øle Stream 116 roach 94, 95, 101 Tange Stream 22 Ølholm Brook 33 Rorippa temperature 16, 33, 45-53, 69, 78, 100- Østerbæk Brook 33 R. amphibia 72 102, 142 Rye Stream 17 Å territory 57, 123 Rødding Stream 124 tourist 96, 130, 147 Århus Stream 31, 32, 34, 50, 115, 116 Trichoptera 75, 79, 80 Åsbæk Brook 116 S trophic index 108, 110, 111 Salmo trout 22, 87, 89-103, 119, 126, 134, 136, S. salar 142 142, 143, 147 S. trutta 87, 92, 142 Tude Stream 17, 34 salmon 93-103, 124, 131, 136, 142, 143, turbidity 46 147 turbulence 56, 61, 63-65, 145 salmonid 24, 102 U sewage 27, 29, 37, 40, 41, 95, 99, 103, 107, 117, 139, 145 Uggerby Stream 17 shear stress 55, 56, 58 Urtica shredders 46 U. dioeca 81 shrimp 87 V Silo S. pallipes 79 Varde Stream 17, 34, 40, 100, 116, 141 Simuliidae 77 Vejle Stream 17 Siphlonurus Viborg Index 118, 120, 121 S. lacustris 116 Vidå Stream 17, 100, 116, 142 Siphonoperla Vindinge Stream 60 S. burmeisteri 116 viscosity 56 Skals Stream 14, 17 Vorgod Stream 34 smolt 96, 97, 99, 102 W Sneum Stream 17, 100 Sparganium 60, 62, 63, 111, 112 water chemistry 27- 29, 53, 100, 142 spates 40, 41 Water Framework Directive 43, 90, species number 106 133-136, 146 species richness 48, 50, 52, 53, 65, 68, water level 35, 57-60, 126, 130 69, 86, 103, 106-109 weed 25, 86, 87, 102, 106, 109, 111, 115, stocking of trout 102, 143 119, 136, 140, 143 stoneﬂ y 77, 82, 118, 121, 141 Wormaldia Stratiotes W. subnigra 116 S. aloides 112 Authors

1 Senior Scientist Annette Baattrup-Pedersen 1 National Environmental Research Institute [email protected] Department of Freshwater Ecology Vejlsøvej 25, PO Box 314 Chief Consultant Christian Dieperink 2 DK-8600 Silkeborg, Denmark [email protected] 2 1 WaterFrame Senior Scientist Nikolai Friberg Præstebakken 3C [email protected] DK-8680 Ry, Denmark Senior Advisor Ruth Grant 1 3 [email protected] Forest and Nature Agency Lindet Forestry Department Biologist Hans Ole Hansen 3 Stensbækvej 29, Arnum [email protected] DK-6510 Gram, Denmark

1 Senior Scientist Carl Christian Hoffmann 4 Institute of Biological Sciences [email protected] University of Aarhus Deputy Director General Torben Moth Iversen 1 Ole Worms Allé, Building 1135 [email protected] DK-8000 Aarhus C, Denmark

Senior Scientist Brian Kronvang 1 5 Bio/consult A/S [email protected] Johannes Ewalds Vej 42-44 DK-8230 Åbyhøj, Denmark Associate Professor Tom Vindbæk Madsen 4 [email protected] 6 Sønderjyllands County Stream Department Head of Monitoring Department Bjarne Moeslund 5 Skelbækvej 2 [email protected] DK-6200 Aabenraa, Denmark Senior Biologist Hans Thiil Nielsen 6 7 [email protected] Freshwater Biological Laboratory Biological Institute Senior Hydrologist Niels Bering Ovesen 1 University of Copenhagen [email protected] Helsingørsgade 51 DK-3400 Hillerød, Denmark Scientist Morten Lauge Pedersen 1 [email protected] 8 Danish Institute for Fisheries Research Assistant Professor Tenna Riis 4 Charlottenlund Castle [email protected] DK-2920 Charlottenlund, Denmark

Full Professor Kaj Sand-Jensen 7 9 Fyns County [email protected] Ørbækvej 100 DK-5220 Odense SØ, Denmark Senior Biologist Jens Skriver 1 [email protected]

Scientiﬁ c Project Manager Ole Vestergaard 8 [email protected]

Section Head Peter Wiberg-Larsen 9 [email protected] Running Waters is about the ecology of Danish streams. It exemplifi es Danish stream research and monitoring over past decades. It is a tale of wasted opportunities and long-term abuse of Danish lowland streams from about 1850 to 1990: Channelisation, intensive management and pollution have in this period deteriorated ecological quality signifi cantly. However, it is also the recent story of improvements accompanying physical restoration of the streams and large-scale water purifi cation.

The state of the environment in Denmark, as in other densely populated countries in Europe, is mostly determined by cultural factors. This is a constraint but also an opportunity. If political willingness and courage exist, streams can recover their physical variability and develop a high diversity of plants, animals and ﬁ sh. We demonstrate that the potential is certainly there.

With many illustrations and a direct language, Danish stream ecologists describe the knowledge we have gained and the excitement of experiencing the varied landscape along meandering streams with their diverse life and potentially rich catch of trout.

Professor Kaj Sand-Jensen, University of Copenhagen, Dr Nikolai Friberg, National Environmental Research Institute, Denmark, and Dr John Murphy, Centre for Ecology and Hydrology, U.K., have edited the book.

National Environmental Research Institute Ministry of the Environment . Denmark ISBN 978-87-7772-929-4