MACROPHYTE-FLOW INTERACTIONS IN AQUATIC PLANTS
By
Julianne Trelenberg
B.Sc., University of Northern British Columbia, 2001

THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF
MASTER OF SCIENCE
in
NATURAL RESOURCES AND ENVIRONMENTAL STUDIES
ENVIRONMENTAL SCIENCE

THE UNIVERSITY OF NORTHERN BRITISH COLUMBIA
October 8, 2004

© Julianne Trelenberg, 2004

1^1

Library and
Archives Canada

Bibliothèque et
Archives Canada

Published Heritage
Branch

Direction du
Patrimoine de l'édition

395 W ellington Street
Ottawa ON K 1A 0N 4
Canada

395, rue W ellington
Ottawa ON K 1A 0N 4
Canada
Your file Votre référence
ISBN: 0-494-04651-1
Our file Notre référence
ISBN: 0-494-04651-1

NOTICE:
The author has granted a non­
exclusive license allowing Library
and Archives Canada to reproduce,
publish, archive, preserve, conserve,
communicate to the public by
telecommunication or on the Internet,
loan, distribute and sell theses
worldwide, for commercial or non­
commercial purposes, in microform,
paper, electronic and/or any other
formats.

AVIS:
L'auteur a accordé une licence non exclusive
permettant à la Bibliothèque et Archives
Canada de reproduire, publier, archiver,
sauvegarder, conserver, transmettre au public
par télécommunication ou par l'Internet, prêter,
distribuer et vendre des thèses partout dans
le monde, à des fins commerciales ou autres,
sur support microforme, papier, électronique
et/ou autres formats.

The author retains copyright
ownership and moral rights in
this thesis. Neither the thesis
nor substantial extracts from it
may be printed or otherwise
reproduced without the author's
permission.

L'auteur conserve la propriété du droit d'auteur
et des droits moraux qui protège cette thèse.
Ni la thèse ni des extraits substantiels de
celle-ci ne doivent être imprimés ou autrement
reproduits sans son autorisation.

In compliance with the Canadian
Privacy Act some supporting
forms may have been removed
from this thesis.

Conformément à la loi canadienne
sur la protection de la vie privée,
quelques formulaires secondaires
ont été enlevés de cette thèse.

While these forms may be included
in the document page count,
their removal does not represent
any loss of content from the
thesis.

Bien que ces formulaires
aient inclus dans la pagination,
il n'y aura aucun contenu manquant.

Canada

11
Abstract
Macrophyte-Flow Interactions
Linear-bladed (Vallisneria americana), whorled (Elodea canadensis), and
dissected (Ceratophyllum demersum) leaf morphologies occur convergently in aquatic
plants. It is likely that fluid dynamics have had an effect on plant morphology, by acting
on exchange processes at low flows and drag forces at higher flows. These species were
examined using flow visualization and digital imagery at five velocities ( 1 - 1 1 cms"') in
a flume. Results indicated that in terms of fluid retention, Elodea and Vallisneria acted
like a cylinder, whereas Ceratophyllum acted like a mesh. Transitional flows occurred at
lower velocities than the cylinder, indicating an increase in local mixing. However, this
was reversed when the intemodal spacing was increased experimentally for Elodea and
Ceratophyllum. This reduction in plant-flow interaction would allow these plants to
remain erect in a relatively low energy environment. These results suggest that the flow
environment has influenced the plasticity of leaf types and the evolution of macrophyte
form in general.

I ll

TABLE OF CONTENTS
Abstract............................................................................................................................ ii
Table of Contents............................................................................................................ iii
List of Tables................................................................................................................... vi
List of Figures..................................................................................................................ix
Chapter 1: Introduction to Thesis
Introduction......................................................................................................... 1
Literature Cited................................................................................................... 5
Chapter 2: Literature Review: Fluid Dynamics and Macrophytes
Introduction......................................................................................................... 8
Fluid Dynamics and Macrophytes
Types of Flow.........................................................................................9
Reynolds Number...................................................................................10
Boundary Layers.....................................................................................II
Shear Stress............................................................................................. 14
Drag......................................................................................................... 15
Froude Number.......................................................................................17
Turbulence Intensity...............................................................................18
Freshwater Macrophytes
Architecture.............................................................................................19
Aquatic versus Terrestrial Plants.......................................................... 26
Adaptations to Flow...............................................................................32
Distribution of Macrophytes................................................................. 33
Flow Around Individual Plants
Photosynthesis........................................................................................ 35
Nutrients..................................................................................................38
Grazing, Predation, and Competition................................................... 39
Flow in Macrophyte Canopies
Effect of Macrophytes on Flow............................................................ 40
Disturbance.............................................................................................43
Velocity Profiles.................................................................................... 44
Sediments................................................................................................45
Methodologies.................................................................................................... 46
Flow Visualization................................................................................. 46
Velocimetry............................................................................................ 48
Approximations and Derived Measurements...................................... 49
Other Measurements.............................................................................. 50
Conclusion...........................................................................................................52
Literature Cited................................................................................................. 54

IV

Chapter 3: The Effect of Macrophyte Morphology (Linear-hladed, Whorled, and
Dissected) on Fluid Dynamics
Abstract................................................................................................................78
Introduction......................................................................................................... 79
Materials and Methods....................................................................................... 82
Flume.......................................................................................................83
Photographs............................................................................................ 83
Top Perspective.........................................................................84
Side Perspective........................................................................ 85
Concentration Curve............................................................................. 85
Image Analysis..........................................................................87
Acoustic Doppler Velocimeter (ADV; SonTek) Measurements
89
Plant Angles and Oscillations............................................................... 89
Statistical Analysis
Effect of Velocity on Dye Coverage andConcentration
90
Effect of Location and Perspective on the Dye Coverage and
Concentration...................................................................... 90
Effect of Morphology on the Rate ofOscillation...................91
Effect of Velocity and Node Position on the Leaf Angles of
Ceratophyllum...................................................................... 91
Effect of Location on Velocity................................................91
Results
Empty Flume..........................................................................................92
Cylinder.................................................................................................. 93
Vallisneria americana...........................................................................94
Elodea canadensis.................................................................................96
Ceratophyllum demersum..................................................................... 97
Comparison of the Three Plant Morphologies and the Physical Model
................................................................................................... 98
Velocity Measurements........................................................... 100
Plant Response to Flow..........................................................................102
Discussion
Macrophyte-Flow Interaction................................................................103
Implications of the Macrophyte-Flow Interaction.............................. 109
Literature Cited...................................................................................................114
Chapter 4: The Effect of Internodal Spacing on Downstream Flow Patterns and
Compliance in Elodea canadensis and Ceratophyllum demersum.
Abstract............................................................................................................... 160
Introduction.........................................................................................................161
Materials and Methods...................................................................................... 163
Statistical Analysis
Effect of Velocity on Dye Coverage andConcentration
165
Effect of Intemodal Spacing on the Rateof Oscillation
165
Effect of Velocity and Node Position on the Leaf Angles of
Ceratophyllum........................................................................165

Results..................................................................................................................166
Cylinder...................................................................................................166
Elodea...................................................................................................... 167
Elodea with the Whorls at Every Second Node Removed................. 167
Elodea with the Whorls at Every Second and Third Node Removed...
168
Ceratophyllum........................................................................................ 168
Ceratophyllum with the Whorls at Every Second Node Removed.... 169
Ceratophyllum with the Whorls at Every Second and Third Node
Removed........................................................................................... 170
Effect of Manipulating Intemodal Spacing
Area of Dye Coverage............................................................... 171
Concentration of Dye................................................................ 173
Angle of Leaf from the Stem...............................
174
Angle of Deflection................................................................... 175
Oscillations.................................................................................175
Comparison of Plant Types.......................................................176
Discussion............................................................................................................178
Literature Cited....................................................................................................183
Chapter 5: Conclusion to Thesis
Conclusion...........................................................................................................218
Literature Cited................................................................................................... 222

VI

List of Tables
Chapter 3: The Effect of Macrophyte Morphology (Linear-hladed, Whorled, and
Dissected) on Fluid Dynamics
Box 3.1: Matlab commands for the image analysis......................................................120
Table 3.1: Shear velocity (u*) determined for each of the velocities used for the
experiments................................................................................................... 121
Table 3.2: Pump settings for isokinetic release of dye at each water velocity

122

Table 3.3: ANCOVA results for the area of dye coverage and theconcentration of
dye for (a) the empty test section, (b) the circular cylinder, (c) Vallisneria,
(d) Elodea, and (e) Ceratophyllum.............................................................. 123
Table 3.4: Post hoc comparisons for the area of dye coverage (Table 3) in the different
domains surrounding different test subjects: (a) empty test section, (b) circular
cylinder, (c) Vallisneria, (d) Elodea, and (e) Ceratophyllum................... 124
Table 3.5: Post hoc comparisons for the concentration of dye (Table 3) in the different
domains surrounding different test subjects: (a) empty test section, (b) circular
cylinder, (c) Vallisneria, (d) Elodea, and (e) Ceratophyllum................... 125
Table 3.6: ANCOVA results for (a) the area covered in dye and (b) the concentration of
dye in that area for the different domains.................................................... 126
Table 3.7: Post hoc comparisons between the subjects for the area covered in dye (Table
6) in the different domains; (a) side upstream, (b) top upstream, (c) side
downstream, and (d) top downstream.........................................................127
Table 3.8: Post hoc comparisons between the subjects for the concentration of dye in the
plume (Table 6) in the different domains; (a) side upstream, (b) top upstream,
(c) side downstream, and (d) top downstream............................................128
Table 3.9: ANCOVA results for velocity in the (a) streamwise component (u) and (b)
vertical component (w) for the test subjects............................................... 129
Table 3.10: Post hoc comparisons for velocity (Table 8a,b) in the (a) streamwise
component (u), and (b) vertical component (w) for the different test
subjects......................................................................................................... 130
Table 3.11: ANCOVA results (a) and post hoc comparison (b) for examining the angle of
deflection and the different test subjects.....................................................131

vil

Table 3.12: ANCOVA results (a) and post hoc comparisions (b) for the frequency of
oscillations between the different test subjects.......................................... 132
Chapter 4: The Effect of Internodal Spacing on Downstream Flow Patterns and
Compliance in Elodea canadensis and Ceratophyllum demersum.
Table 4.1: ANCOVA results for (a) the area covered in dye and (b) the concentration of
dye in that area for the different domains for all of the test subjects
combined....................................................................................................... 186
Table 4.2: ANCOVA results for the area of dye coverage and the concentration of dye for
(a) the circular cylinder, (b) Elodea, (c) Elodea with the leaves at every
second node removed, and (d) Elodea with the leaves at every second and
third node removed.......................................................................................187
Table 4.3: Post hoc comparisons between Elodea and its associated manipulations for the
area covered in dye in the (a) side perspective upstream, (b) in the top
perspective upstream, (c) in the side perspective downstream, and (d) in the
top perspective downstream........................................................................ 188
Table 4.4: ANCOVA results for the area of dye coverage and the concentration of dye for
(a) the circular cylinder, (b) Ceratophyllum, (c) Ceratophyllum with the leaves
at every second node removed, and (d) Ceratophyllum with the leaves at every
second and third node removed................................................................... 189
Table 4.5: Post hoc comparisons between Ceratophyllum and its associated
manipulations for the area covered in dye in the (a) side perspective upstream,
(b) in the top perspective upstream, (c) in the side perspective downstream,
and (d) in the top perspective downstream................................................. 190
Table 4.6: Post hoc comparisons between Elodea and its associated manipulations for the
concentration of dye in the (a) side perspective upstream, (b) in the top
perspective upstream, (c) in the side perspective downstream, and (d) in the
top perspective downstream.........................................................................191
Table 4.7: Post hoc comparisons between Ceratophyllum and its associated
manipulations for the concentration of dye in the (a) side perspective
upstream, (b) in the top perspective upstream, (c) in the side perspective
downstream, and (d) in the top perspective downstream......................... 192
Table 4.8: (a) ANCOVA and (b) post hoc comparisons for the angle of the
Ceratophyllum leaves (and the manipulations thereof) from the stem in the up
and downstream position..............................................................................193

V lll

Table 4.9: (a) ANCOVA and (b) post hoc comparisons for the angle of deflection
between Elodea and Ceratophyllum and there associated manipulations
........................................................................................................................ 194
Table 4.10: (a) ANCOVA and (b) post hoc comparisons for the frequency of oscillations
for Elodea, Ceratophyllum, and there associated manipulations............. 195

List of Figures
Chapter 2; Literature Review; Fluid Dynamics and Macrophytes
Figure 2.1: a) Laminar flow as seen by the smooth dye streakline, b) transitional flow as
seen by the inereasing waviness of the streakline, and c) turbulent flow as seen by
the choatic nature of the streakline....................................................................63
Figure 2.2: The transition to turbulent flow of a fluid as it passes a circular cylinder seen
in cross section at different Reynolds numbers: (a) 1, (b) 10, (c) 13, (d) 26, (e)105,
and (f) 150........................................................................................................... 64
Figure 2.3: The benthic boundary layer concept.A velocity profileis presented form the
surface (U = 0) to the free stream (U = Uo»; from Ackerman 1989).............. 65
Figure 2.4: A rectangular block of fluid is deformed by shear stress, so two faces become
parallelograms....................................................................................................66
Figure 2.5: The four basic habitats of freshwater macrophytes: (A, B) free-floating; (C)
submerged; (D, F) floating-leaved; and (E, F) em ergent............................... 67
Figure 2.6: Examples of caulescent macrophytes a) Elodea canadensis and b)
Myriophyllum spicatum; rosette macrophytes c) Isoetes spp. and d) Vallisneria
americana; and a thalloid macrophyte e) Tristicha altemifolia.....................68
Figure 2.7: Different arrangements of leaves around plant stems: a) opposite, b) alternate,
c) whorled, and d) rosulate.................................................................................69
Figure 2.8: Morphologies of entire leaves ranging from linear to sagittate (cordate)
..............................................................................................................................70
Figure 2.9: An example of a petiole-like freshwater macrophyte (Marsilea vestitd)
..............................................................................................................................71
Figure 2.10: Different leaf margins that occur within entire leaf morphologies

72

Figure 2.11: Fenestration occurs when small areas of interstitial tissue are lacking at
m aturity.............................................................................................................. 73
Figure 2.12: Dissected leaves occur when the leaf is subdivided (e.g., Ceratophyllum)
..............................................................................................................................74
Figure 2.13: The leaves represent the transition from submerged (left) to aerial leaf (right)
morphology observed in the vertically ascending stems of Proserpinaca palustris.
..............................................................................................................................75

Figure 2.14: Typical growth forms of freshwater macrophytes in slow to very fast flows
in shallow and deeper flowing w aters..............................................................76
Figure 2.15: Two examples of heterophylly: (a) Cabomba caroliniana has dissected
submerged leaves and entire floating leaves, and (b) Potamogeton natans has
linear submerged leaves and ovate floating leaves.......................................... 77
Chapter 3: The Effect of Macrophyte Morphology (Linear-hladed, Whorled, and
Dissected) on Fluid Dynamics
Figure 3.1: Lighting and dye injector set up for the photos.........................................133
Figure 3.2: Velocity profiles in the empty test section at: (a) 0.013 m s'\ (b) 0.020 m s'\
(c) 0.050 m s'\ (d) 0.084 ms"\ and (e) 0.11 ms'^.........................................134
Figure 3.3: Diagram of the positioning device attached to the side of the flume with the
cylinder in place for the top perspective........................................................ 135
Figure 3.4: Cuvette set up for creating the concentration curve.................................. 136
Figure 3.5: Original concentration curve (pixel value vs. dye concentration) for the (a)
side and (b) top perspectives...........................................................................137
Figure 3.6: (a) log (pixel value) versus log (dye concentration) in the side perspective for
0.01-5% dye, (b) pixel value versus dye concentration in the side perspective for
10-100% dye, (c) log (pixel value) versus log (dye concentration) in the top
perspective for 0.01-5% dye, and (d) pixel value versus dye concentration in the
top perspective for 10-100% dye....................................................................138
Figure 3.7: Mock photo to test analysis method in MATLAB.................................... 139
Figure 3.8: Grid used to take ADV measurements in the test section (120 cm downstream
of the flow straighteners).................................................................................140
Figure 3.9: Flow visualization in the empty test section using fluoroscein dye in the side
(left) and the top (right) perspectives............................................................. 141
Figure 3.10: Regression of area of dye coverage (a-d) and concentration of dye (e-h)
versus velocity or Reynolds number (based on the hydraulic diameter) in the
different domains for the empty test section..................................................142
Figure 3.11: Flow visualization around a 0.7 cm diameter circular cylinder using
fluoroscein dye in the side (left) and the top (right) perspectives................143

XI

Figure 3.12: Regression of area of dye coverage (a-d) and concentration of dye (e-h)
versus velocity or Reynolds number (based on the diameter) in the different
domains for the circular cylinder....................................................................144
Figure 3.13: Flow visualization around Vallisneria americana using fluoroscein dye in
the side (left) and the top (right) perspectives............................................... 145
Figure 3.14: Regression of area of dye coverage (a-d) and concentration of dye (e-h)
versus velocity or stem Reynolds number (based on the plant diameter) in the
different domains for Vallisneria americana................................................ 146
Figure 3.15: Flow visualization around Elodea canadensis using fluoroscein dye in the
side (left) and the top (right) perspectives..................................................... 147
Figure 3.16: Regression of area of dye coverage (a-d) and concentration of dye (e-h)
versus velocity or stem Reynolds number (based on the plant diameter) in the
different domains for Elodea canadensis...................................................... 148
Figure 3.17: Flow visualization around Ceratophyllum demersum using fluoroscein dye
in the side (left) and the top (right) perspectives...........................................149
Figure 3.18: Regression of area of dye coverage (a-d) and concentration of dye (e-h)
versus velocity or stem Reynolds number (based on the diameter) in the
different domains for Ceratophyllum demersum...........................................150
Figure 3.19: Comparison of the fluid dynamic conditions around plants and a physical
model (a) The velocity at which the transition from laminar to turbulent occurs
for the different plants; (b) Rate of change in the area of dye coverage with
increasing velocity for the different test subjects; and (c) The rate of change in
the concentration of dye with increasing velocity for the different test
subjects..............................................................................................................151
Figure 3.20: Contours of the velocity field measured 3 cm from the flume bottom in a 2
cm X 2 cm grid in an empty flume..................................................................152
Figure 3.21: Contours of the velocity field measured 3 cm from the flume bottom in a 2
cm X 2 cm grid (3 cm in the vicinity of the cylinder) around a circular
cylinder............................................................................................................. 153
Figure 3.22: Contours of the velocity field measured 3 cm from the flume bottom in a 2
cm X 2 cm grid (3 cm in the vicinity of the plant) around Vallisneria
americana......................................................................................................... 154
Figure 3.23: Contours of the velocity field measured 3 cm from the flume bottom in a 2
cm X 2 cm grid (3 cm in the vicinity of the plant) around Elodea
canadensis........................................................................................................ 155

X ll

Figure 3.24: Contours of the velocity field measured 3 cm from the flume bottom in a 2
cm X 2 cm grid (3 cm in the vicinity of the plant) around Ceratophyllum
demersum..........................................................................................................156
Figure 3.25: Regressions for the angle of deflection for (a) the circular cylinder and the
different plant species (b) Vallisneria, (c) Elodea, and (d) Ceratophyllum.
Regressions for the number of oscillations per second with increasing velocity
for (e) the circular cylinder and the different plant species (f) Vallisneria, (g)
Elodea, and (h) Ceratophyllum.......................................................................157
Figure 3.26: (a) Rate of change in the angle of deflection with change in velocity for the
different test subjects; and (b) the rate of change in the number of oscillations
per second with change in velocity for the different test subjects............... 158
Figure 3.27: Flow patterns around a circular cylinder at increasing Reynolds numbers: (a)
Re < 1, the flow is laminar; (b) Re > 4, attached, recirculating eddies; (c) Re >
80, alternatively shed vortices (von Karman vortex street); (d) Re > 200, flow
within vortices becomes chaotic; and (e) Re > 5000, wake is completely
turbulent (after Kundu, 1990)........................................................................ 159
Chapter 4: The Effect of Internodal Spacing on Downstream Flow Patterns and
Compliance in Elodea canadensis and Ceratophyllum demersum.
Figure 4.1: Flow visualization around a 0.7 cm diameter circular cylinder using
fluoroscein dye in the side (left) and the top (right) perspectives................196
Figure 4.2: Regression of area of dye coverage (a-d) and concentration of dye (e-h)
versus velocity or Reynolds number (based on the diameter) in the different
domains for the circular cylinder....................................................................197
Figure 4.3: Flow visualization around Elodea canadensis using fluoroscein dye in the
side (left) and the top (right) perspectives..................................................... 198
Figure 4.4: Regression of area of dye coverage (a-d) and concentration of dye (e-h)
versus velocity or stem Reynolds number (based on the plant diameter) in the
different domains for Elodea canadensis...................................................... 199
Figure 4.5: Flow visualization around Elodea with the whorl at every second node
removed using fluoroscein dye in the side (left) and the top (right)
perspectives...................................................................................................... 200
Figure 4.6: Regression of area of dye coverage (a-d) and concentration of dye (e-h)
versus velocity or stem Reynolds number (based on the plant diameter) in the
different domains for Elodea canadensis with every second node removed...
201

Xlll

Figure 4.7: Flow visualization around Elodea with the whorls at every second and third
node removed using fluoroscein dye in the side (left) and the top (right)
perspectives..................................................................................................... 202
Figure 4.8: Regression of area of dye coverage (a-d) and concentration of dye (e-h)
versus velocity or stem Reynolds number (based on the plant diameter) in the
different domains for Elodea canadensis with every second and third node
removed...........................................................................................................203
Figure 4.9: Flow visualization around Ceratophyllum demersum using fluoroscein dye in
the side (left) and the top (right) perspectives............................................. 204
Figure 4.10: Regression of area of dye coverage (a-d) and concentration of dye (e-h)
versus velocity or stem Reynolds number (based on the diameter) in the
different domains for Ceratophyllum demersum......................................... 205
Figure 4.11: Regression analysis for the leaf angle from the stem of Ceratophyllum in the
upstream and downstream position from the plant stem........................... 206
Figure 4.12: Flow visualization around Ceratophyllum with the whorl at every second
node removed using fluoroscein dye in the side (left) and the top (right)
perspectives...................................................................................................... 207
Figure 4.13: Regression of area of dye coverage (a-d) and concentration of dye (e-h)
versus velocity or stem Reynolds number (based on the diameter) in the
different domains for Ceratophyllum demersum with every second node
removed............................................................................................................ 208
Figure 4.14: Regression analysis for the leaf angle from the stem of the Ceratophyllum
with every second node removed in the upstream and downstream position from
the plant stem................................................................................................... 209
Figure 4.15: Flow visualization around Ceratophyllum with the whorl at every second
and third node removed using fluoroscein dye in the side (left) and the top
(right) perspectives.......................................................................................... 210
Figure 4.16: Regression of area of dye coverage (a-d) and concentration of dye (e-h)
versus velocity or stem Reynolds number (based on the diameter) in the
different domains for Ceratophyllum demersum with every second and third
node removed...................................................................................................211
Figure 4.17: Regression analysis for the leaf angle from the stem of the Ceratophyllum
with every second and third node removed in the upstream and downstream
position from the plant stem........................................................................... 212

XIV

Figure 4.18: The rate of change in the angle of the leaf from the stem with change in
velocity for Ceratophyllum and its associated manipulations for the leaves (a)
facing the approaching flow and (b) in the downstream flow..................... 213
Figure 4.19: Regression results for the angle of deflection versus velocity for (a) the
circular cylinder, (b) Elodea, (c) Elodea 2“'^, (d) Elodea 3'^, (e) Ceratophyllum,
(f) Ceratophyllum 2”'^, and (g) Ceratophyllum 3’^'^........................................ 214
Figure 4.20: Regression results for the frequency of oscillation with increasing velocity
for (a) the circular cylinder, (b) Elodea, (c) Elodea 2"'^, (d) Elodea 3^^, (e)
Ceratophyllum, (f) Ceratophyllum 2"^, and (g) Ceratophyllum 3"^.............215
Figure 4.21: Comparison of the fluid dynamic conditions around plants and a physical
model (a) The velocity at which the transition from laminar to turbulent occurs
for the different plants; (b) Rate of change in the area of dye coverage with
change in velocity for the different test subjects; and (c) The rate of change in
the concentration of dye with change in velocity for the different test
subjects............................................................................................................. 216
Figure 4.22: (a) Rate of change in the angle of deflection with change in velocity for the
different test subjects; and (b) the rate of change in the number of oscillations
per second with change in velocity for the different test subjects...............217

Chapter 1; Introduction to Thesis
The natural flow regime of rivers is dynamic, and influences the population and
community structure of aquatic organisms, with abiotic processes dominating at higher
velocities (i.e., flow regime: Poff et al., 1997; Bernez et al., 2004), and biotic processes
dominating at slower velocities (i.e., grazing and competition: Abrams, 1980; Poff and
Ward, 1989). Among these organisms, freshwater macrophytes have the ability to
modify the physiochemical aspects of their microenvironment by slowing water flow,
which, in turn, leads to the trapping of sediments and increase in organic content, and a
change in the water chemistry profiles (i.e., due to photosynthesis and nutrient uptake;
Chambers et al., 1999). Consequently, freshwater macrophytes can be considered
ecosystem engineers in that they create and alter habitats for other organisms by affecting
physical resources such as light, sediments, and water flow (Posey et al., 1993; Alper,
1998; Crooks, 2002; Wright et al., 2004). At the larger scale, macrophytes can alter the
geomorphology of an area by affecting erosion and sedimentation rates thus affecting the
geometry of water channels, and the hydrology of an area by affecting the surface flow
patterns (Gordon, 1998).
As indicated above, macrophytes affect their environment but, the environment
also affects them. As a result, many aquatic plants are highly plastic morphologically,
and similar morphological patterns occur in unrelated species. With the exception of
fenestration (a morphology found in the tropics in a select number of species (i.e.,
Aponogeton; Sculthorpe, 1967), three main morphological patterns occur in freshwater
macrophytes: linear-bladed (e.g., Vallisneria americana)', simple whorled (e.g., Elodea
canadensis)', and dissected (e.g., Ceratophyllum demersum). Variations in these patterns

occur when plants of the same species (or even the same plant) exposed to different
environmental conditions exhibit heterophylly or foliar plasticity. For example,
submerged leaves tend to be highly dissected, whereas aerial leaves tend to be much
thicker with smoother margins (Sculthorpe, 1967; Wells and Pigliucci, 2000).
Additionally, the length of the intemodal space in whorled species can change depending
on environmental conditions, with the space being longer for plants in sheltered sites
(Idestam-Almquist and Kautsky, 1995), or in areas of low light (Cronin and Lodge,
2003).
In terms of flow environment, plants are exposed to diffusional stresses at slower
velocities (Hurd et al., 1997) and to mechanical stresses at higher velocities (Schutten and
Davy, 2000). Diffusional stresses occur because at slower velocities there is a thick
diffusional boundary layer, that limits the supply of nutrients to the leaf surface layer
(Cheer and Koehl, 1987; Koehl, 1996; Hurd et al., 1997). In order overcome these
limitations, some macrophytes have small roughness elements on their leaves (Abelson et
al., 1993; Hurd, 2000) that change the conditions in the boundary layer from laminar to
turbulent, thus increasing the amount of local mixing, and consequently nutrient uptake
by the plants. At faster velocities, productivity can be limited by tissue damage and
dislodgment (Koch, 1993; Stewart and Carpenter, 2003), self-shading (Koehl and
Alberte, 1988; Niinemets and Fleck, 2002), and the inhibition of enzymes (Koch, 1994).
However, the central issue to living in high flow conditions is the reduction of drag,
which is accomplished by the plant becoming smaller in stature and/or compliant with the
flow (Ennos, 1999; Sand-Jensen, 2003), through a change in shape and orientation
(Schutten and Davy, 2000).

Given the common morphological patterns found in macrophytes and the
importance of fluid flow to productivity and ecology it would be appropriate to examine
whether plants with different leaf types (linear-bladed, whorled and dissected) interact
with the flow in the same manner. Furthermore, it is important to examine whether the
intemodal spacing effects the flow patterns due to flow separation (Coutanceau and
Defaye, 1991) and their compliance under different velocities. This examination should
also include a comparison to a physical model of a circular cylinder, which is often used
as an approximation for aquatic plants (e.g. Nepf, 1999). Specifically, the major features
of the plants (e.g., entire leaves, dissected leaves, and rigidity) can be contrasted with the
model.
The question of the macrophyte-flow interaction is addressed in Chapter 2
through a literature review of: (1) basic fluid dynamic concepts; (2) the architecture of
aquatic plants and how they differ from those in terrestrial systems; (3) adaptation of
plants to a flow environment at both the individual and canopy scale, and the implications
on distribution, photosynthesis, nutrient uptake, and grazing, predation, and competition;
and (4) some of the methodologies used the characterize flow conditions. This
background leads to the examination of the null hypothesis that plants with different leaf
types (linear-bladed, whorled and dissected) affect the downstream fluid in the same
manner in Chapter 3. This is done by taking short video sequences of Fluoroscein dye
moving past the different plant morphologies, and conducting image analysis to
determine the area of dye coverage and the concentration of dye in specified areas around
the three different plant morphologies and a physical model. Similar techniques are used
in Chapter 4 to examine the null hypothesis that intemodal spacing has no effect on the

flow patterns generated by the plants and their compliance under different velocities.
Together, these chapters combine to address the issue of whether the morphological
patterns seen in freshwater macrophytes have been influenced by fluid dynamics factors.

Literature Cited
Abelson, A., T. Miloh, and Y. Loya. 1993. Flow patterns induced by substrata and body
morphologies of benthic organisms, and their roles in determining availability of food
particles. Limnology and Oceanography. 38(6): 1116-1124.
Abrams, P. 1980. Some comments on measuring niche overlap. Ecology. 61(l):44-49.
Alper, J. 1998. Ecosystem ‘engineers’ shape habitats for other species. Science.
280(5367): 1195-1196.
Bemez, L, H. Daniel, J. Haury, and M.T. Ferreira. 2004. Combined effects of
environmental factors and regulation on macrophyte vegetation along three rivers in
western France. River Research and Applications. 20:43-59.
Chambers, P.A., R E. DeWreede, E.A. Irlandi, and H. Vandermeulen. 1999. Management
issues in aquatic macrophyte ecology: a Canadian perspective. Canadian Journal o f
Botany. 77:471-487.
Cheer, A.Y.L, and M.A.R. Koehl. 1987. Paddles and rakes: fluid flow through bristled
appendages of small organisms. Journal o f Theoretical Biology. 129:17-39
Coutanceau, M., and J. Defaye. 1991. Circular cylinder wake configurations: A flow
visulization survey. Applied Mechanics Reviews. 44(6):255-305.
Crooks, J.A. 2002. Characterizing ecosystem-level consequences of biological invasions:
the role of ecosystem engineers. Oikos. 97:153-166.
Cronin, G., and D.M. Lodge. 2003. Effects of light and nutrient availability on the
growth, allocation, carbon/nitrogen balance, phenolic chemistry, and resistance to
herbivory of two freshwater macrophytes. Oecologia. 137:32-41.
Gordon, D R. 1998. Effects of invasive, non-indigenous plant species on ecosystem
processes: Lesson from Florida. Ecological Applications. 8(4):975-989.
Ennos, A.R. 1999. The aerodynamics and hydrodynamics of plants. The Journal o f
Experimental Biology. 202:3281-3284.
Hurd, C.L. 2000. Water motion, marine macroalgal physiology, and production. Journal
ofPhycology. 36:453-472.
Hurd, C.L, C.L Stevens, B.E. Laval, G.A. Lawrence, and P.J. Harrison. 1997.
, Visualization of seawater flow around morphologically distinct forms of the giant kelp
Macrocystis integrifolia from wave-sheltered and exposed sites. Limnology and
Oceanography. 42(1): 156-163.

Idestam-Almquist, J., and L. Kautsky. 1995. Plastic responses in morphology of
Potamogeton pectinatus L. to sediment and above-sediment conditions at two sites in
the northern Baltic proper. Aquatic Botany. 52:205-216.
Koch, E.W. 1993. The effect of water flow on photosynthetic processes of the alga Ulva
lactucah. Hydrobiologia. 260/261:457-462.
Koch, E.W. 1994. Hydrodynamics, diffusion-boundary layers and photosynthesis of the
seagrass Thalassia testudinum and Cymodocea nodosa. Marine Biology. 119:767776.
Koehl, M.A.R. 1996. When does morphology matter? Annual Review o f Ecological
Systems. 27:501-542.
Koehl, M.A.R., and R.S. Alberte. 1988. Flow, flapping, and photosynthesis of
Nereocystis luetkeana: a functional comparison of undulate and flat blade
morphologies. Marine Biology. 99: 435-444.
Nepf, H.M. 1999. Drag, turbulence, and diffusion in flow through emergent vegetation.
Water Resources Research. 35(2): 479-489.
Niinemets, U., and S. Fleck. 2002. Petiole mechanics, leaf inclination, morphology, and
investment in support in relation to light availability in the canopy of Liriodendron
tulipifera. Oecologia. 132:21-33.
Poff, N.L., J.D. Allan, M B. Bain, J R. Kar, K.L. Prestgaard, B.D. Ritcher, R E. Sparks,
and I.e . Stromberg. 1997. The natural flow regime. Bioscience. 47(11): 769-782.
Poff, N.L., and J.V. Ward. 1989. Implications of streamflow variability and predictability
for lotie community structure: A regional analysis of streamflow patterns. Canadian
Journal o f Fisheries and Aquatic Sciences. 46:1805-1817.
Posey, M.H., C. Wigand, and J.C. Stevenson. 1993. Effects of an introduced aquatic
plant, Hydrilla verticillata, on benthic communities in the upper Chesapeake Bay.
Estuarine, Coastal and Shelf Science. 37:539-555.
Sand-Jensen, K. 2003. Drag and reconfiguration of freshwater macrophytes. Freshwater
Biology. 48:271-283.
Schutten, J., and A.J. Davy. 2000. Predicting the hydraulic forces on submerged
macrophytes from current velocity, biomass and morphology. Oecologia. 123:445452.
Sculthorpe, CD . 1967. The Biology o f Aquatic Vascular Plants. Edward Arnold, London
(Reprinted 1985 Koeltz Scientific, Konigstein), 610 pp.

Stewart, H.L., and R.C. Carpenter. 2003. The effects of morphology and water flow on
photosynthesis of marine macroalgae. Ecology. 84(11):2999-3012.
Wells, C.L., and M. Pigliucci. 2000. Adaptive phenotypic plasticity: the case of
heterophylly in aquatic plants. Perspectives in Plant Ecology, Evolution and
Systematics. 3(1):1-18.
Wright, J.P., W.S.C. Gurney, and C.G. Jones. 2004. Patch dynamics in a landscape
modifies by ecosystem engineers. Oikos. 105:336-348.

Literature Review

8

Fluid Dynamics and Macrophytes
Introduction
Freshwater aquatic plants (macrophytes) are important components of lake and
river ecosystems (Biggs, 1996). They are considered ecosystem engineers because they
create, and alter habitat for aquatic organisms by: (1) slowing or modifying water
velocities; (2) altering oxygen regimes; (3) blocking light; (4) trapping detritus; (5)
providing refuges from predators and disturbance; (6) providing oviposition sites; and (7)
increasing the amount of attachment surface (Biggs, 1996; Wright et al., 2004). In
addition, emergent vegetation around the edges of rivers and lakes help to reduce
shoreline erosion by dampening the effect of wave energy (Kalff, 2002). Conversely,
dense beds of submerged macrophytes in riverine systems can impair discharge, thus
increasing the potential for flooding, as well as interfering with the use of water craft
(Kalff, 2002). Macrophyte communities also have important ecological effects whereby
they may increase the pH of poorly buffered waters, reduce dissolved oxygen levels, alter
stream courses and bed roughness, shift invertebrate communities from free stone
dwelling taxa to small burrowing taxa, and degrade the appearance of the stream (Biggs,
1996). These positive and negative aspects of freshwater macrophytes speak to the
importance of understanding vegetated hydraulic systems; especially since macrophytes
increase the complexity of underwater landscapes (Chambers et al., 1999).
There are several critical components of the flow regime that regulate ecological
processes in river ecosystems including: the magnitude; frequency; duration; timing; rate
of change of hydraulic conditions in a set location; and development of the mainstream
flow (Poff et al., 1997). There are also factors that affect the smaller scale velocities

Literature Review

9

within macrophyte beds, which include: velocity over unvegetated substrate; biomass of
macrophytes at a specific location; biomass of macrophytes immediately upstream of the
location; and the depth of the water at that site (Gregg and Rose, 1982). At the river and
the macrophyte scale, the physical structure of the riverine environment can be defined
largely by physical processes, especially those related to the movement of water and
sediment (Poff et al., 1997). The following chapter will examine fluid dynamic issues on
the smaller scale, specifically, the flow patterns around individual macrophytes and
through canopies. These smaller scale fluid dynamics are important because they speak
directly to plant productivity through nutrient uptake (Borchardt, 1994), photosynthesis
(Dennison, 1987), and sediment capture (Gregg and Rose, 1982). In addition, a basic
understanding of fluid dynamics must be presented before one can examine how flow
influences freshwater plant communities, and how flow is in turn influenced by
individual plants. Patterns of water movement can be examined both qualitatively and
quantitatively in order to characterize the macrophyte-flow interaction. It will become
evident that the macrophyte-flow interaction is important for the autoecology of plants
and the ecology of aquatic systems.
Fluid Dynamics and Macrophytes
Types of Flow
The most logical place to begin a discussion of macrophyte-flow interaction is to
look at the flow conditions, specifically, laminar, transitional, and turbulent (Figure 2.1).
In laminar flow, it is assumed that all fluid particles move nearly parallel to each other in
a smooth path. The large- and small-scale movements of the fluid are the same, at least
down to the level at which molecular diffusion becomes the dominant mode of transport

Literature Review

10

(Vogel, 1994). Transition flow, where the particle paths begin to deviate from a parallel
path, occurs between laminar and turbulent flow, with the transition often occurring
quickly (Vogel, 1994). Finally, turbulent flow occurs when individual fluid particles
move in a chaotic fashion, even if the fluid as a whole appears to be traveling uniformly
in one direction. In other words, intense small-scale motion in all directions is
superimposed on the mean velocity (Vogel, 1994). Virtually all flows of aquatic interest
are turbulent at large spatial scales (Nowell and Jumars, 1984), and as such, laminar and
transitional flows are generally not applicable to natural systems except on very small
scales (e.g., around individual leaves and within canopies; Ackerman and Okubo, 1993;
Ackerman, 2002).
Reynolds Number
Reynolds number ( Re = —

; where u = velocity, 1 = characteristic length, and v

V

= the kinematic viscosity [the ratio of dynamic viscosity to density]) provides an
indication of the flow conditions (Figure 2.1, 2.2), as it is a ratio of inertial to viscous
forces (Vogel, 1994). The situation where viscous flow dominates (e.g.. Re «

1;

creeping flow) is not dealt with in much detail here, rather the more typically flow
conditions where inertia is relevant (e.g.. Re > 1) are presented. If the Reynolds number
is low (e.g. small, slowly moving structures), the situation is relatively viscous, and the
flow is smooth and orderly (e.g.. Figure 2.2a). At high Reynolds numbers (e.g. large,
rapidly moving structures), inertial forces dominate, and the flow is turbulent (Koehl,
1996). In natural rivers the transition to turbulent flow often occurs around Re = 10^ 10"^ using the hydraulic diameter of the channel, but the actual value is dependent on the
roughness of the bed relative to the water depth (Carling, 1992).

Literature Review

11

Boundary Lavers
Velocity gradients occur near boundaries when the fluid comes in contact with a
solid object due to the ‘no-slip’ condition at the boundary (Figure 2.3). This causes the
velocity of the flow to decline toward the boundary (Thompson and Trioan, 1997)
because the velocity of a fluid at the surface of a solid is actually zero. In general, a
higher Re implies a thinner boundary layer but a higher shear rate (as will be defined in
the next section). Similarly, the lower the Re, the thicker the boundary layer and the
lower the shear rate in the boundary layer (Cheer and Koehl, 1987). The thickness of the
boundary layer (5) around a cylinder shaped object under laminar flow conditions can be
approximated by the formula: Ô ~ 5x/^Re^

, where 5 is a constant (which may range

from 4.65-5.84), x is the downstream distance, and Rcx is the local Reynolds number
taken x distance from the leading edge (Vogel, 1994; Zhang and Malmqvist, 1997).
Additionally, shear is a consequence of the dissipation of momentum and energy, so
another way of viewing the boundary layer is as the region near a surface where the
action of viscosity produces a large loss of total pressure head (Vogel, 1994).
Drag is another important concept related to boundary layers. Drag is the
hydrodynamic force that pulls a body in the direction opposite to fluid movement and in
the case of boundary layers, the force that slows the flow (Koehl, 1996). At low Re, drag
is due to skin friction, which is the viscous resistance of the fluid in the boundary layer
around the body being sheared as the fluid moves past the body, so a greater wetted area
leads to higher skin friction. At high Re drag is also due to skin friction but is dominated
by pressure drag, which is the pressure difference across the body due to the fluid
dynamic separation or the formation of a wake on the downstream side of the body

Literature Review

12

(Vogel, 1987; Koehl, 1996). These concepts are important when considering boundary
layers because differences in velocity can control the thickness of the boundary layer
close to the plant surface and as both velocity and plant size increase so does the skin
friction and drag, which effects the plant community (Biggs, 1996).
An example of the importance of boundary layers in macrophyte ecology can be
seen in plants with an undulate or bullate blade morphology where the boundary layer
thickness along the curved body changes due to changing pressure gradients. Hurd et al.
(1997) demonstrated that when water moves up the rising portion of an undulation of a
macrophyte blade, the velocity increases and the boundary layer thickness decreases.
Conversely, as water moves down the descending portion of an undulation, the velocity
decreases and the thickness of the boundary layer increases rapidly. Macroalgae in waveexposed versus sheltered sites have different morphologies (smooth vs. undulate) because
of contrasting demands of diffusional and mechanical stress. At low flows characteristic
of sheltered environment, seaweed productivity may be limited by diffusion due to the
development of a thick boundary layer that reduces the transport of essential inorganic
nutrients to the blade surface (Hurd et al., 1996). Undulations are interpreted to be the
plant’s response to water motion that increase local mixing. Conversely at high flows
characteristic of wave-exposed environments, drag can damage or detach seaweeds from
the bottom (Hurd et al., 1997). Smooth blades cause little water motion, specifically
separation, allowing for less drag, which is an advantage in highly energetic
environments because it prevents the plants from being broken, damaged, or uprooted.
Undulate blades cause an increase in the water motion around them thus increasing the

Literature Review

13

drag, which is advantageous in low energy environments because this added movement
limits self-shading and increases mixing.
Changes in boundary layers are important to the productivity of aquatic plants for
a number of reasons. First, turbulence caused by the interaction of flow with leaves
greatly decreases the time that nutrients are exchanged relative to molecular diffusion
alone. Second, where flow is laminar, as in lentic systems and at very small scales in
other systems, boundary layer thickness depends on the velocity, and the size and
orientation of the macrophyte blade (Losee and Wetzel, 1993; Crossley et al., 2002).
Within dense macrophyte beds, thick boundary layers are the result of reduced turbulence
caused by the hydroelastic response of plants (Ackerman and Okubo, 1993). In other
words, viscous forces dominate over turbulent forces at a small distance from the plant or
sediment surface. Conceptually, a layer of water sticks to plant and sediment surfaces
and does not partake in the watver circulation (Kalff, 2002).
The concept of the roughness Reynolds number (Re* = u*k/v where u* = shear
velocity =^frTp , x = shear stress, p = density, k = height of roughness elements, and v =
kinematic viscosity), provides a better description of the flow situation close to the
bottom, than the Re of the mean flow (Davis and Barmuta, 1989). This is because, the
height of the roughness elements relative to the thickness of the viscous sublayer of the
boundary layer (see below) determines the flow conditions near the bed (Davis and
Barmuta, 1989). When the height of the roughness element is smaller than the thickness
of the viscous sublayer, flow conditions are considered to be smooth turbulent, whereas if
the height of the roughness is greater than the thickness of the viscous sublayer, the flow
conditions are considered rough turbulent (Davis and Barmuta, 1989). Smooth turbulent

Literature Review

14

flows can only occur in smooth flow conditions, such that occurs over fine sediments, or
adjacent to streamlined submerged macrophytes (Davis and Barmuta, 1989).
Shear Stress
Turbulent shear stress (x), which is the resistive force caused by the no-slip
condition, can be determined from a log-linear regression of the logarithmic portions of
velocity profiles (or viscous sublayer; Figure 2.3; Ackerman, 1997; Ackerman and
Hoover, 2001), providing enough measurements are taken to achieve reasonable
confidence limits (Nowell and Jumars, 1984). However, it must be mentioned that nonlogarithmic layers may occur in strongly accelerating or decelerating flow such as in river
reaches which narrow or widen rapidly (Hoover and Ackerman, in Press). The
logarithmic profile may also be deformed by spatial variation in the mixture of the
bottom roughness, which would include the presence of large rocks (Carling, 1992).
Shear stress is perhaps the most meaningful measure of flow as far as benthic organisms
are concerned because it is the shearing force of water rather than velocity that is likely to
erode or dislodge organisms (Lancaster and Hildrew, 1993). Shear forces are also
important because they set a limit to the size of fine particles that can settle to the
sediment surface. For example, at faster flows fine particles will not settle, creating a
coarse substrata (Sand-Jensen, 1998). As such, the evaluation of the bed shear stress is
important, because it is necessary not only for normalizing turbulence characteristics but
also for assessing issues regarding sediments (Nezu and Onitsuka, 2001).
Laminar shear stress ( t =

, where p is the dynamic viscosity, z is the

height (or depth) and du/dz is the local velocity gradient) is important in momentum
transfer to the leaf surface. Since the viscosity of the fluid is directly related to the

Literature Review

15

thickness of the boundary layer (i.e., how rapidly layers of fluid move with respect to
each other), the presence of the no-slip condition denotes an area where shear stress
exists (Vogel, 1988). Hence, the more viscous a fluid the larger the velocity gradient of
the boundary layer and, the less viscous a fluid the steeper the velocity gradient of the
boundary layer (Vogel, 1988). This is important because the process of nutrient
exchange to a plant’s surface is enhanced by steeper boundary layers, which increase the
rate of momentum transfer and thus the replenishment of the site of exchange.
Drag
Drag is a property of fluid dynamics that is of interest when considering the
productivity of macrophytes. As indicated above there are two types of drag, skin
friction and pressure drag (Vogel, 1988; 1994). In other words, the drag force on an
object in a moving fluid depends on the velocity of the fluid, the size, and roughness of
the object. Additionally, for a non-spherical object, the shape and orientation in relation
to the flow are also important (Schutten and Davy, 2000). The drag on small organisms
is mainly due to skin friction, whereas, for larger (and more upright) organisms pressure
drag tends to be much greater (Biggs and Thomsen, 1995). An example of this occurs
when plant stems bend downstream in response to faster flow, so that the frontal area is
reduced, lowering pressure drag, whereas long ribbon-shaped leafs can fold against the
stem, reducing skin friction (Ennos, 1999). Consequently, pressure drag will be high on
dense plant communities because of the pressure drop from upstream to downstream.
This occurs because there is an area of slow flow at the end of the canopy due to the flow
separation around the canopy (Vogel, 1994). Conversely, pressure drag will be lower in
open plant stands due to the additional drag exerted by the plants that reduce the mean

Literature Review

16

flow within vegetated regions relative to unvegetated ones, thus making the energy
distribution even throughout the flow (Ackerman and Okubo, 1993; Vogel, 1994; Nepf,
1999). A consequence of this velocity/pressure range is that more individual plants will
have access to nutrients in open vs. dense canopies, whereas plants in dense canopies will
have more protection from drag. For example, drag and acceleration forces may limit
productivity by dislodging macrophytes from the substratum (Hurd et al., 1996).
At some point the momentum acquired by a macrophyte must be stopped because
the plant is anchored to the substratum. A large, flexible structure can move with the
flow for a time, thereby reducing drag and lift (a force perpendicular to the direction of
flow), but this can only occur while acquiring momentum (Denny et al., 1997).
Accelerating a body in a fluid involves a force with three additive components: (1) drag;
(2) the force needed to accelerate the mass of the body forward; and (3) the force needed
to accelerate the displaced mass of fluid backward. The additive forces, which depend
upon the shape of the object, the volume of the body, and the density of the surrounding
fluid, is commonly called the acceleration reaction. In other words, if an object is
accelerated in one direction, an equal volume of fluid must be accelerated in the other
direction. As such, to accelerate an object, a force proportional to the object’s volume
and to the density of the medium must be applied (Vogel, 1994). This concept is relevant
to aquatic systems because when a current flows past a plant, the plant resists
acceleration thus causing a displacement of the fluid. For example, if the plant is large
enough, it can simply comply as water passes by reducing the relative velocity and
acceleration between the plant and the surrounding water and, thereby, reducing the

Literature Review

17

hydrodynamic forces (Denny et al., 1997). These concepts are especially important in
intertidal systems involving kelps (Dawes, 1981).
Froude number
Ü

The Froude number, (Fr =

where u = mean velocity, g = acceleration due to

gravity, and D = water depth), is the ratio of inertial to gravitational forces that describes
the surface condition in a flowing system (i.e., the ratio of the inertia of the flow to the
speed of a wave). The inertial force causes plant movement (i.e., compliance with the
flow) or divert it from a straight course at constant speed (i.e., oscillations; Vogel, 1988).
For conditions where Fr < 1 the flow is designated as sub-critical or tranquil flow, when
Fr = 1 flow is critical, and for Fr > 1 flow is super-critical. Super-critical flow is
characterized by broken, white water and is described as shooting or streaming flow
(Davis and Barmuta, 1989). Nezu and Onitsuka (2001) showed that in biological
systems, the maximum value of vorticity increases with an increase in Froude number.
They also noted that the strength of the secondary currents increased with an increase of
the Froude number. Secondary currents vary in flow direction from the mean flow,
leading to what one might consider very large scale turbulence intensity. In other words,
Froude number is relevant to macrophytes because it also characterizes the mean flow
(Davis and Barmuta, 1989), and it describes the distribution of momentum in the main­
stream velocity (Kundu, 1990). Correspondingly, drag increases more quickly with
velocity at the surface than in the water column (Vogel, 1994). Consequently, floating
and emergent leaved plants exist in still or very slow moving water, whereas submerged
plants, which are more adapted to comply to drag conditions exist in faster moving
environments (Dawson, 1988).

Literature Review

18

Turbulence Intensity
Mean velocity gives a measure of the average unidirectional flux of water past
macrophyte blades, whereas turbulence intensity (i.e.,w^ / if *100, where Urms is the root
mean square of the velocity) provides an indication of the velocity fluctuations that may
lead to local mixing (Koehl and Alberte, 1988). Vegetation not only affects the mean
velocity, but also affects the turbulence intensity and the resultant transport of nutrients.
The conversion of mean flow to turbulent flow within stem wakes increases the
turbulence intensity, and because wake turbulence is generated at the stem scale, the
dominant turbulent length scale is reduced, relative to unvegetated, open-channel
conditions (Ackerman and Okubo, 1993; Nepf, 1999). The combination of reduced
velocity and reduced eddy-scale should reduce the in-canopy turbulent diffusion relative
to unvegetated regions (Ackerman and Okubo, 1993; Nepf, 1999).
The eddy viscosity (the turbulent transfer of momentum that characterizes the
transport and dissipation of energy in the smaller-scale flow), which is determined by the
properties of the turbulence, characterizes mean momentum transfer by turbulence
fluctuations (Mathieu and Scott, 2000). As such, it is an essential property of mass
transfer to aquatic plants, since eddy viscosity is not constant within the boundary layer
(Hinze, 1987) or the plant canopy (Ackerman and Okubo, 1993). For example, it has
been shown that turbulence intensity is enhanced by the presence of a mesh, yet the
diffusivity is diminished because the eddy scale is reduced (Nepf, 1999). That being
said, it is clear that turbulent flows modify the rate at which materials and energy are
exchanged between a macrophyte and its immediate environment (Hart et al., 1996).

Literature Review

19

The random motion of eddies is an efficient way of transferring mass and
momentum, because each eddy carries with it the properties of the fluid contained within
it (Hurd, 2000). Mass transfer is an important concept in aquatic systems, whereby
nutrients and wastes are transported towards and away from organisms. Any nutrient
uptake by the macrophyte blade generates a nutrient depleted region adjacent the blade
surface, called the concentration boundary-layer (Stevens and Hurd, 1997). If the
timescale for significant changes in the viscous sublayer is much less than the timescale
for molecular transport, then it is possible for there to be reduced net advection (Stevens
and Hurd, 1997) that can be related to a thick diffusive boundary layer and ultimately to
diffusional stress. In turbulent flow, mass shifts around in directions other than that of
the overall flow (Vogel, 1994). For this reason, the thinner the diffusive boundary layer
in the water flowing along a blade, the greater the rate of mass transfer to the blade
surface (Koehl and Alberte, 1988). Thus, fluid dynamics can define the environment
where freshwater macrophytes are productive.
Freshwater Macrophytes
Architecture
The term freshwater macrophyte is broad, and used to encompass emergent,
floating, floating-leaved and submerged plants that include representatives of
Angiosperms, Pteridophytes, Bryophytes and large algae (Dawson, 1988). Plant
architecture is also a broad term, but it is defined here to include only the
photosynthetically active part of the plant. It is important to understand the function of
plant architecture, and to determine whether it is a result of common ancestry, or whether
it is due to similar environmental pressures. The uncertainty of the origin of plant

Literature Review

20

architecture arises from the observation that freshwater macrophytes in different families
have similar morphologies, whereas species in the same family can have different
architectures. This may be because plants generally acelimate to changing environmental
conditions through morphological rather than physiological means (Wells and Pigliucci,
2000; Santamaria, 2002). Heterophylly (foliage plasticity) occurs in some species of
freshwater macrophytes whereby the plant has more than one kind of leaf structure
depending on whether they are under, on, or above the water surface. Heteroblastic
development is a form of heterophylly whereby juvenile leaves have a different
morphology than the adult leaves regardless of the environment (Sculthorpe, 1967).
The classification of organisms and the evolutionary relationships among them
(systematics) is determined using comparisons of the evolved similarities (e.g.,
synapomorphies) in plant characteristics such as genetics and morphology (Raven et al.,
1999). However, this is sometimes difficult in aquatic plants, because they exhibit a wide
range of phenotypic plasticity (the ability of an individual organism to alter its
morphology and physiology in response to environmental conditions; Idestam-Almquist
and Kautsky, 1995; Wells and Pigluicci, 2000; Santamaria, 2002) as well as similarity of
form (Sculthorpe, 1967). Phytogeny is also a difficult concept to apply to aquatic plants
because they represent a functional group. For instance, it is known that aquatic
Pteridophytes (i.e., ferns) and Angiosperms (i.e., herbs) evolved from terrestrial ancestors
polyphyletically (Sculthorpe, 1967), thus providing examples of convergent evolution.
However, heterophylly occurs across distant taxa, which suggests that unrelated plants
have adapted in a similar fashion to the aquatic environment (Wells and Pigliucci, 2000);

Literature Review

21

adding to the argum ent that structural similarities occur due to adaptation to a com mon

environment.
Freshwater macrophytes vary in morphology at different scales, ranging from leaf
shape to growth form. There are four basic growth forms of vascular macrophytes: (1)
emergent; (2) floating-leaved; (3) submerged; and (4) free-floating (Figure 2.5).
Emergent plants (e.g., Sparganium) tend to be mainly rhizomatous (having a horizontal,
usually underground stem that often sends out roots and shoots from its nodes) or
cormous (having a short thick solid food-storing underground stem) and are perennial.
Heterophylly can be experienced in emergent plants due to the submerged or floating
leaves preceding the aerial leaves in development (Sculthorpe, 1967). In addition,
emergent macrophytes have aerial reproductive structures. Floating-leaved plants (e.g.,
Nuphar and Nympheae) can either have leaves attached to long flexible petioles that are
rhizomatous or cormous or they can have compliant stems that rise through the water and
produce floating leaves on relatively short petioles and be stoloniferous (stems growing
horizontally above the ground and producing roots and shoots at the nodes). The
reproductive structures are either floating or aerial, and consequently, heterophylly can be
experienced since the submerged leaves precede the floating leaves in development
(Schulthrope, 1967). Submerged macrophytes can be divided into three main types: (1)
caulescent; (2) rosette; and (3) thalloid (Figure 2.6). Caulescent plants (ones with a welldeveloped aboveground stem), can be with or without rhizomes, and consist of long
flexible stems that root from the nodes (e.g., Elodea and Myriophyllum). Rosette types
(e.g., Isoetes and Vallisneria) are often stoloniferous, and have radical leaves arising from
a condensed, often tuberous rhizome. Lastly, plants where the body is reduced to a

Literature Review

22

cylindrical or flattened form, often bearing secondary branches (e.g., Podostemaceae) are
of the thalloid type (Sculthorpe, 1967). Submerged leaves tend to be filiform, ribbon­
shaped, fenestrated, or finely divided (Sculthorpe, 1967). Finally, free-floating plants
include small surface floating plants (e.g.. Lemma and Wolffia), as well as plants that are
submerged with few or no roots (e.g., Ceratophyllum; Sculthorpe, 1967). Considering
these different growth forms, submerged plants generally occur in deeper water, floating
leaved communities occur closer to the shore, and emergent plants occur in reed-swamp
communities, although this zonation can, and does overlap (Schulthorpe, 1967).
In freshwater macrophytes, there are two main patterns of growth, the first being
an abbreviated axis producing a rosette of radical leaves (e.g., Isoetes spp.), the second
being an elongated flexuous stem which is covered in leaves and rooted from its nodes
(e.g., Myriophyllum; Sculthrope, 1967). There are several ways that leaves can be
arranged around the stems of these macrophytes. For example, the leaves of caulescent
macrophytes may be: (1) alternate (e.g., Potamogeton)', (2) paired and opposite (e.g.,
Cabomba); whorled (e.g., Hippuris and most species of Elodea and Myriophyllum-,
Figure 2.7); or (3) pseudo-whorled whereby the leaves actually occur alternately, but
several condensed intemodes occur between two successive long intemodes (e.g.,
Myriophyllum heterophyllum) (Sculthorpe, 1967).
There are three main leaf forms in freshwater macrophytes, these being entire,
fenestrated, and dissected. Entire leaves are generally thin and translucent, the shape
varying from awl-like and linear to ovate and sagittate (or cordate; Figure 2.8). Entire
leaves can also be petiole-like with sheathing at the base and upper limb with the leaf
being either wide or narrow where it attaches to the petiole (Figure 2.9). Within the

Literature Review

23

entire morphology, many of the broad leaf types either have undulate margins (Figure
2.10) or bullate surfaces (Sculthorpe, 1967). Undulate margins along with fenestration
occur in the leaves of the genus Aponogeton, which are entire and narrowly lanceoate. In
this case, the margin is slightly undulate to tightly crinkled, and the leaf is generally
twisted because the marginal regions grow faster than then median regions. The
fenestration in Aponogeton madagascariensis is observed by the small areas of interstitial
tissue that are lacking at maturity, sometimes to the extent that the lamina is thoroughly
perforated leaving a skeleton like structure of veins (Figure 2.11; Sculthorpe, 1967).
Thirdly, dissected leaves (e.g., Ceratophyllum) occur when the leaf is split into many free
segments radiating from the petiole. The leaves of Ceratophyllum are quite rigid, 1-4
times forked, and have two rows of teeth along the segments at the tip of the leaf (Figure
2.12; Sculthorpe, 1967). Alternatively, the leaves of Myriophyllum spicatum are also
dissected, but instead of being radial in structure, the segments grow alternately from a
central vein, giving a feathery appearance.
As mentioned previously, leaf variations can be found: (1) across species; (2)
among the population of the same species; and (3) between the leaves of a single plant
(heterophylly; Wells and Pigliucci, 2000). For example, heterophylly occurs in
Prosperpinaca palustris as a response to light quantity (Figure 2.13). When the leaves
are submerged, or experience only eight hours of light a day, the leaves are highly
dissected. Whereas the aerial leaves, or the leaves that experience sixteen hours of light a
day are entire lanceolate leaves with serrated margins (Wells and Pigliucci, 2000).
Heterophylly can also be induced by light quality as in Hippuris vulgaris whereby

Literature Review

24

changes in red:far red signal the transition from submerged to aerial leaf morphology
(Wells and Pigliucci, 2000).
Hormones also play a role in changes to plant morphology. It has been shown
that plants exposed to abscisic acid produce aerial leaf morphology while still submerged.
Abscisic acid is the drought stress hormone in terrestrial plants that regulates plants to
changes in water availability. Gibberellic acid encourages the production of leaves
characteristic of submerged plants while still in an aerial environment, but it is considered
to be a response to photoperiod rather than water availability (Wells and Pigliucci, 2000).
Freshwater macrophytes have different morphological features depending on their
environment. For example, lake sites with high wind exposure are usually dominated by
small compact (rosette-like) species with stiff leaves arising from the base of the stem,
and a strong root system. In contrast, species with long, slender highly flexible stems
usually dominate shallow river channels or shorelines of low turbulence, which allow
rooting and prevent damage to the stems. Macrophytes with floating leaves, such as
Nuphar, are restricted to areas of little or no flow (Dawson, 1988; Kalff, 2002). Erect
and canopy-forming species with apical growth, commonly dominate in deeper waters
under relatively sheltered conditions (Figure 2.14; Kalff, 2002). To expand upon this
concept, it has been shown that plant species occurring in highly energetic and disturbed
environments are predicted to be more robust; i.e., shorter in height (Stewart and
Carpenter, 2003), limited in lateral spread, shorter lived, have a larger proportion of
annual production devoted to seeds, as well as having fewer storage structures (BarratSegretain, 2001). Conversely, competitive species that occur in more predictable habitats
are predicted to be more delicate (i.e., taller in height; Stewart and Carpenter, 2003),

Literature Review

25

extensive in lateral spread, longer lived, have a smaller proportion of production for
annual reproduction, as well as having important vegetative storage structures (BarratSegretain, 2001). To further contrast the difference between low and fast flow, it has
been observed that plants from fast-moving streams tend to have leaves with low
branching angles (i.e., present a smooth surface to the flow), whereas, plants from slowmoving streams tend to have higher branching angles, and the leaves stick out more into
the flow (i.e., present a rough surface to the flow; Ennos, 1999). This increase in angle
also increases the drag, but results in the production of small vortices behind the leaves
that would increase the supply of carbon dioxide and other nutrients (Ennos, 1999).
It is reasonable to suggest that the physical interaction of plants with the water
flowing past them governs and selects for suitable plant forms and species. Thus the
plants’ size, form and stand structure, together with the strength of its stems and the
security of its method of anchoring to the substratum, are major factors in plant survival
(Dawson, 1988). Another factor that may be useful in the description of architecture is
the leaf-area index, which provides an estimate of the photosynthetically active surface in
the macrophyte community, as well as the amount of substrate available that specific
community (Den Hartog, 1982).
Several factors influence near-bed velocity and sediment composition within
macrophyte beds in lowland streams, which are also closely related to differences in
macrophyte morphology (Barrat-Segretain, 2001). Anderson and Charters (1982)
suggested that water motion affects marine macroalgae through the transport of mass,
momentum, and energy from the main body of the fluid through the boundary layer to the
plant’s surface. The transport rates depend strongly on whether the flow in the boundary

Literature Review

26

layer is laminar or turbulent, because the transfer rates in turbulent flow with turbulent
diffusion are orders of magnitudes greater than those in laminar flow with rates
determined by molecular diffusion alone. Plants occurring within or near the boundary
layer or those that create their own unstirred layer through dense canopies tolerate the
least hydrodynamic stress, whereas plants that are less densely branched and stronger are
predicted to break at higher velocities than highly branched filaments and tufts (Sheath
and Hambrook, 1988), thus demonstrating that where aquatic plants of different
morphologies grow is dependent mainly on the fluid dynamics in that area.
Aquatic versus Terrestrial Plants
There are several important environmental (i.e., chemical and physical) factors
that affect plants in aquatic systems. The first of these comparisons is of a mechanical
and structural nature, which examines how plants support themselves in air compared to
water. Secondly, the physiology of gas exchange is important when dealing with the
transport of important nutrients given the different diffusivities of oxygen (0 %) and
carbon dioxide (CO 2) in air versus water. Another factor to consider is photon capture
(discussed in a later section) which, relates directly to the productivity of plants and how
light is attenuated in aquatic versus terrestrial environments. There is also the issue of
mechanical damage caused by hydraulic stress, which will differ between aquatic and
terrestrial systems. Finally, ecological factors such as pollination and diaspore dispersal
differ given the lack of animal agents in aquatic systems.
One of the main differences between living in water versus air is the density of
the medium. For example, the density of air varies directly with temperature, pressure,
and altitude, with a variation of about 13%. Conversely, changes in the density of water

Literature Review

27

are small (0.8%) over the biologically relevant range of temperatures (Denny, 1993).
However, since water weighs much more then air, a small difference in density can have
large implications for aquatic organisms (Denny, 1993). Another major difference
between the density of air and water is that water density does not respond in a linear
fashion to temperature; rather, water goes through a transition phase whereby ice is less
dense than water, and so it floats (Denny, 1993). Similarly, if the density of any object is
greater than the surrounding fluid, the object experiences a downward force proportional
to its volume (negatively buoyant). Conversely, if the density of an object is less than the
surrounding fluid, there is a net upward force (positively buoyant). Consequently, since
the density of water is so close to the body density of plants, the effective or excess
density (pd - Pf where Pd = density of the body, and pf = density of the fluid) is very
sensitive to small changes in the density of the fluid. Comparatively, because the density
of air is so small, it has little effect on the effective density of terrestrial plants (Denny,
1993).
One consequence of the differences in density discussed above, is that terrestrial
and aquatic plants have different means of supporting themselves. Specifically,
terrestrial plants are rigid with a dominant axis, whereas aquatic plants do not possess a
dominant axis because they are partially supported by the water itself. The support given
to submerged macrophytes is also provided by the buoyancy provided by lancunate
tissues (air filled spaces that run along the leaf in the mesophyll), which reduces the need
for mechanical strength and rigidity (Sculthorpe, 1967). For this reason, aquatic stems do
not need to support the weight of leaves, as is the case in terrestrial environments.
Additionally, given the differences in density discussed above, and the fact that aquatic

Literature Review

28

plants have low bulk densities, aquatic plants tend to be neutrally or positively (due to
lacunae) buoyant. In contrast, terrestrial plants are roughly a thousand times denser than
air and experience the compressing effects of gravity (Niklas, 1997).
The density differences between air and water translate to differences in
kinematic viscosity, which is the ratio of the molecular or dynamic viscosity to density.
In addition, there are differences in the molecular diffusivity of gases related to
photosynthesis. For example, even though the concentration of carbon dioxide (CO2) is
the same in water and air between of 10°C and 20°C, a terrestrial leaf receives CO 2 at a
higher rate than that of an aquatic leaf due to greater diffusivity in air. Thus, aquatic
plants must overcome diffusional stresses by living in areas of mixing (i.e., moving
water), or they must have smaller or more dissected leaves (i.e., to increase surface area:
volume) than terrestrial plants to maintain an adequate delivery rate of CO2. There is also
a possibility that leaf edges influence the fluid and thus induce turbulence (Coutanceau
and Defaye, 1991). Another difference between aquatic and terrestrial environments is
related to the dissociation of CO 2 in water (Denny, 1993). An equilibrium is reached in
aquatic systems whereby when CO2 dissolves in water, much of it combines chemically
with its surroundings to produce carbonic acid (H2CO3), which then dissociated into a
hydrogen ion and a bicarbonate ion (HCO3 ), which is often used by aquatic plants as
their carbon source for photosynthesis. The final ion in the equilibrium, carbonate (COs^"
), is not used by plants. Carbon dioxide has a lower molecular diffusivity in freshwater
than in air, consequently as indicated above, plants living in air can acquire CO 2 more
efficiently than their aquatic counterparts (Niklas, 1997). That being said, land plants and

Literature Review

29

aquatic mosses are only able to use carbon dioxide (CO2), but many aquatic seed plants
and algae are also able to use HCO3' (Sculthorpe, 1967).
The differences in the diffusivity of oxygen and carbon dioxide in air and water,
may relate to the morphological differences between submerged macrophytes and the
aerial form of the same species. This heterophylly is the presence on a single individual
of two or more distinct types of leaf. For example, Proserpinaca palustris in the aerial
form, has leaves with toothed margins that are smaller, broader, thicker, and fewer in
number than those on submerged stems, which are highly dissected (Figure 2.13;
Schulthorpe, 1967). In extreme examples of heterophylly, the leaf types are different in
three aspects, habit, shape and anatomy, and a single plant may bear submerged, floating,
and aerial leaves (Sculthorpe, 1967). There are many aquatic plants (e.g. Cabomba,
Ceratophyllum, Potamogeton, and Sparganium) that change the morphology of their
leaves when the shoot apex reaches the water surface and emerges into the aerial
environment. This change usually consists of the submerged form being either linear or
highly dissected, to an aerial form that is lanceolate or a more entire leaf form
respectively (Figure 2.15; Sculthorpe, 1967).
On a smaller scale, the terrestrial plants have an external layer of waxy material
(cuticle) that reduces the rate of water loss, the internalization of exchange surfaces such
as superficial pores or stomata, and the development of conducting tissues (Niklas, 1992).
These adaptations exist because terrestrial plants are exposed to the drier atmosphere,
whereas aquatic plants are continuously covered in water (Niklas, 1997). It is important
to note that aquatic plants do not possess stomata and as such nutrient exchange can
occur over the entire surface (Sculthorpe, 1967). Consequently, the limiting state for

Literature Review

30

uptake in aquatic plants is diffusional stress, rather than stomatal resistance in terrestrial
plants (Eugster and Hesterberg, 1996; Vesala, 1998).
Terrestrial and aquatic plants also experience different mechanical stresses. One
way that mechanical stresses are increased in both environments is through an increase in
height, which is a strategy used by plants to maximize light capture for photosynthesis
(Ennos, 1999; Strand and Weisner, 2001). As such, at increasing velocities, a strategy of
minimizing mechanical stress in aerial environments is for leaves to reconfigure into a
conical or cylindrical shape, and become more tightly rolled as wind speed increases. A
similar phenomenon occurs with clusters of leaves, which form bundles that become
tighter as wind speed increases (Vogel, 1994). In aquatic systems ribbon-shaped plants
(e.g. Vallisneria spp.) become more compliant with the flow (streamlined) due to their
lack of strengthening tissue, which reduces the hydrodynamic drag (at higher flows) and
thus the mechanical damage (Sculthorpe, 1967). However, in order to colonize faster
moving environments, the length of the leaf must be small since drag is a function of
surface area (Vogel, 1994; Wilson et al., 2003). Highly dissected leaves are also able to
reduce mechanical stresses by the segments grouping together thereby reducing the
surface area exposed to the approaching flow (Vogel, 1994). Hence, in order to reduce
drag, plants must either become compliant with the flow, or reduce the size of the
individual surfaces facing the flow. Thus, it is clear that aquatic plants have adapted to
these hydrodynamic constraints, however, the mechanisms are as yet not understood.
There are limitations to pollination in submerged aquatic plants due to a number
of factors including the lack of animal agents in aquatic environments. Although the
vegetative features of aquatic plants show modifications of morphology and anatomy in

Literature Review

31

response to the aquatic environment, the reproductive structures are generally similar to
those of related terrestrial plants in both general organization and microscopic structure
(Sculthorpe, 1967; Ackerman, 1995). The flowers of the majority of aquatic angiosperms
are adapted to aerial life, with either insects or wind as pollinating agents (Ackerman
1995, 2000). Pollen can also be dispersed by water with the pollen being carried by
wind, water currents, or surface tension (Ackerman, 2000). Truly submerged pollination
also occurs whereby the drifting pollen, which are individually filiform or adherent in
long chains, become entangled in the typically filiform, feathery, or peltate stigmas
(Sculthorpe, 1967). Since there are no known aquatic organisms that transfer pollen
among aquatically pollinated plants, pollen is transferred on or in the water by wind,
gravity, and water currents (Ackerman, 2000).
There are several dispersal mechanisms for macrophytes. One of these is by
animals and water birds, which are the main agents in the short and potentially long range
dispersal of aquatic plants. Additionally, many submerged species spread by runners or
by vigorous growth in their rhizomes (Philbrick and Les, 1996). Another means of
dispersing, as with Isoetes spp., is to produce plantlets (young plant with a developing
rhizome, adventitious roots, and two leaves) instead of sporangia (Sculthorpe, 1967).
However, floods and normal river currents are probably the most powerful agents in
distributing vegetative propagules. Whole plants of all life forms, rooted rhizomes and
tubers, vegetative fragments, and tuiions (small specialized shoots), are often tom away
and carried for kilometers (Sculthorpe, 1967); thus demonstrating how aquatic plants may
be dispersed over large distances.

Literature Review

32

Adaptations to Flow
Macrophytes must be able to support themselves in a variety of freshwater
environments. It has been noted by Stevens and Hurd (1997) that under almost any field
conditions the pressure gradients at leading edges of macrophyte leaves will be sufficient
to cause separation in the flow. Separation points occur when the flow stops moving up
the surface of the plant, and starts moving in the downstream direction (Vogel, 1994),
where vortices are shed. Thus, with vorticity being generated in the wakes of the leaf
edges, the vortex streets from the leaf wakes combine to produce the turbulence in the
flow within and downstream of the plant body (Anderson and Charters, 1982).
Accordingly, one of the consequences of increased branching angle is to increase
vorticity, and with it the amount of nutrient and light getting to the plant (Denny and
Robertson, 2002), whereas the function of decreased branching angle is to increase the
amount of compliance and thus lower the drag that the plant experiences (Ennos, 1999;
Speck, 2003).
Macrophytes have various strategies to overcome high fluid forces including:
greater structural strength; greater flexibility; growth in thin layers; and by the restriction
of their growth to non-critical seasons of water flow (Dawson, 1988). The more
compliant plants adopt a more streamlined and compact form with increasing velocity
(until the velocity is so high that they become damaged or dislodged). Less compliant
plants have a low initial resistance and can tolerate small increases in velocity by having
leaves at the water surface, but hydraulic resistance increases rapidly as leaves bend into
the water and become submerged (Dawson, 1988). However, most aquatic plant species
have flexible stems allowing them to bend into the flow, thus reducing the surface area

Literature Review

33

directly exposed to the approaching flow (Sand-Jensen, 2003). Fortunately, compliance
can occur without reducing the plant’s capacity for photosynthesis or nutrient uptake, and
without negatively affecting productivity (Gerard, 1987), as will be discussed in the
photosynthesis section. Flexibility also determines which sort of flow habitat is most
mechanically stressful; for rigid organisms, waves produce larger forces than do
unidirectional currents of the same peak velocity, whereas the opposite is true for flexible
organisms (Koehl, 1996). Plants that grow in energetic environments tend to be small in
stature and limited in lateral spread, whereas plants that grow in predictable habitats tend
to be higher in stature and extensive in lateral spread (Barrat-Segretain, 2001). However,
most giant kelps are both long and flexible while living in an energetic environment, and
this morphology may be able to reduce the drag experienced by these plants (Denny et
al., 1997). It has also been found that species with fewer branches break at higher current
velocities as compared to highly branched species (Sheath and Hambrook, 1988). Plant
architecture is clearly important as the shoot area defines the domain of the interaction
with the water current, and the ability of the plant to uptake nutrients (Schutten and Davy,
2000). The relative magnitudes of the stem’s compliance and the anchorage strength
determine whether an individual plant remains in position, breaks, or is uprooted. More
specifically, compliance is the ability of the shoot structure to deform with increasing
current velocity, thus reducing its roughness and frontal area (Schutten and Davy, 2000),
and increasing the likelihood of the plant remaining anchored.
Distribution of Macrophytes
The vegetation of freshwater environments is defined by the effect and
interactions of a single physical factor, water flow, which determines plant-form, growth-

Literature Review

34

controlling factors, and habitat (Dawson, 1988). Dawson (1988) reported that at high
flows, aquatic vegetation may be confined to the margins of the channel and to islands
where emergent vegetation can be in direct competition with terrestrial vegetation. Very
low flows allow the development of a vegetation characteristic of still water (e.g.
emergent and floating plants; Dawson, 1988). Other factors to consider in regards to
macrophyte distribution are that low flows may be accompanied by reduced oxygen
concentrations, increased temperature, and desiccation (i.e., small ponds and streams can
dry up due to extreme climatic variations; Lancaster and Hildrew, 1993), all of which can
influence stream communities. Conversely, unusually high flows are often accompanied
by increased velocity and hydraulic forces (i.e., drag and shear stress) on the stream bed
(Lancaster and Hildrew, 1993). At intermediate flow regimes, the physical and chemical
factors can control growth, interact with aquatic plants to regulate seasonal biomass
(Dawson, 1988). For this reason, the plant community at a site reflects the balance
achieved between the physiochemical environment and the plant’s tolerance, adaptation
to, or the modification of these conditions by the plants’ presence (Dawson, 1988). Other
factors influencing the distribution and growth of freshwater macrophytes are known to
include turbulence, seasonal temperature patterns, surface and interstitial water
chemistry, sediment composition, light, and other biota (White and Hendricks, 2000).
Establishment of well-developed macrophyte patches, however, appear to be related
primarily to constant or predictable discharge, turbidity, and canopy cover. Additionally,
most macrophytes will grow where favorable combinations of substrate type, water
depth, surface velocity, nutrient availability, and other environmental factors, occur
(White and Hendricks, 2000).

Literature Review

35

Flow Around Individual Plants
Photosynthesis
Flow around individual plants directly influences the productivity of the plant as
well as the plant canopy. For example, macroalgae rely on the movement of water for the
delivery of nutrients to their surface, as such, an increase in velocity increases
photosynthesis and thus, productivity (Stewart and Carpenter, 2003). However, many
abiotic and biotic factors control net macrophyte productivity including: photon flux
density and spectral composition (Dennison, 1987; Falkowski and Raven, 1997); nutrient
availability (Borchardt, 1994); temperature (Davison, 1991); water motion (Wheeler,
1988); inter- and intra-specific competition for space and resources (Creed et al., 1996);
and rates of herbivory (Leonard et al., 1998; reviewed in Hurd, 2000). Although, it is
generally agreed that light intensity is the limiting factor in determining the maximum
depth at which plants will occur in a given body of water. The depth where respiration
equals photosynthesis for a particular plant is known as the compensation depth; plants
cannot exist below this depth (Riemer, 1993). Due to the limitations surrounding photon
capture in aquatic systems, freshwater macrophytes must maximize photosynthesis. As
indicated by Hurd and Stevens (1997) flapping is a technique used by marine algae in
order to enhance photon capture by reducing self-shading. Schutten and Davy (2000)
also noted that at low current velocities, the spreading of leaves increases the area for
light interception. However, when exposed to faster currents, the plant receives less
efficient light interceptions due to the resulting compliance, which reduces resistance and
minimizes mechanical damage (Koehl and Alberte, 1988). Another strategy used by
plants to maximize photon capture under poor light conditions is to have longer branches

Literature Review

36

and taller plants (associated with a decreased number of branches and below-ground
biomass); this is also a physical response to flow (Strand and Weisner, 2001). The most
basic morphological response however, is to increase the size of the leaf surface and the
number of photosynthetic cells on that surface (Rascio, 2002).
Gas exchange and light capture are essential to the productivity of plants, making
heterophylly a prime example of the functional explanation for plant morphology.
Accordingly, an increase in surface area: volume is a technique used to maximize light
capture (Gerard and Mann, 1979; Gutschick, 1999), and to increase the flux of nutrients
available to the plant (Wheeler, 1988; reviewed in Hurd, 2000). Fine-scale
morphological features might act as roughness elements generating transitional or
turbulent flow at the plant surface at low mainstream velocities (Koch, 1994; reviewed in
Hurd, 2000). Additionally, the hairs of some macroalgae may increase the deposition of
particulate material and create a still region at the thallus surface in which extracelluar
enzymes can act (Koch, 1993).
Photon capture differs between aquatic and terrestrial environments due to the
different spectral characteristics of light in these two systems. This is related to the fact
that water attenuates the intensity of sunlight and preferentially absorbs the red
wavelengths of light. In contrast, air neither attenuates nor alters the spectral properties
of light (Niklas, 1997), although forest canopies have been found to reduce the red to farred ratio (the ratio between transmitted light in the red band (655-665 nm) to far-red light
(725-735 nm; Lieffers et al., 1999). Light can penetrate only a relatively small distance
(as compared to the atmosphere) into an aqueous medium because water absorbs light
more strongly than air. For example, light traveling through 50 m of air, will lose almost

Literature Review

37

none of its intensity, but only 78% of incident blue light will still present, and virtually
none of the incident red light when light travels through 50 m of water (Denny, 1993). In
addition to this, turbidity decreases water transparency and thus increases light
attenuation (James et al., 2004), which leads to a decrease in plant biomass (Christian and
Sheng, 2003). For this reason, aquatic plants living a few centimeters below the water
surface can experience significantly lower light intensities shifted in favour of blue
wavelengths than plants living in air (Niklas, 1997). In addition to the problem of the
reception of necessary intensities of appropriate wavelengths of light, a photosynthetic
organism under water must also acquire dissolved carbon (Sculthorpe, 1967). Structural
modifications of submerged leaves to increase their photosynthetic efficiency include: an
extremely thin cuticle; thin leaves; and the presence of chloroplasts in the epidermis (as
opposed to being found only in the mesophyll as with land plants; Sculthorpe, 1967).
Leaf shape, the length of intemodes, and the pattern of leaf arrangement are all important
in defining the total amount of direct light a shoot can receive (Niklas, 1992). For
example, it has been documented that in terrestrial plants, leaves that are more erect tend
to have better light-use efficiency (Gutschick, 1999). However, in aquatic systems, a
canopy with horizontal leaves is more successful in low light conditions, where leaf
angles become more erect with increasing irradiance (Niinemets and Fleck, 2002). In
terms of the length of the intemodes, low light levels increase this length as a way to
extend leaves higher in the water column (Cronin and Lodge, 2003). Moreover, because
chlorophyll a (the most common photosynthetic pigment) absorbs strongly at a
wavelength of 680-710 nm (red light; Kirk, 1994), it is a relatively ineffective means for

Literature Review

38

gathering light at depth (Denny, 1993). Interestingly, the diversity of accessory pigments
found in the algae, are not seen in freshwater macrophytes (Dawes, 1981).
Nuritents
Barrat-Segretain (2001) has shown that one modification to poor nutrient
conditions is for the shoot to root ratio to increase with decreasing sediment fertility in
turbulent flow. The shoot to root ratio varies greatly among species and growth forms
with bottom dwelling angiosperms having the lowest ratio, and taller erect and canopyproducing angiosperm species generating the highest ratios. Angiosperms growing in
nutrient-poor environments have disproportionately large root systems and a reduced
shoot to root ratio (Kalff, 2002) to aid in nutrient uptake as well as with anchorage.
Additionally, plants at sheltered sites with thick diffusive boundary layers, increase their
below-ground biomass, compared to plants at wave-exposed sites, probably in order to
supply the plant with nutrients through root uptake (Strand and Weisner, 2001). It is
known that rooted submerged macrophytes can take up inorganic nutrients from both the
sediments and from the water column, but the rooted macrophytes generally obtain most
of their phosphorus and nitrogen from the sediments (White and Hendricks, 2000; Kalff,
2002). This is important because nutrients, especially nitrogen and phosphorus, usually
control the rates of growth after disturbance (Biggs, 1996). Consequently, the maximum
biomass that macrophytes achieve at a particular site, is the result of a balance between
the conditions available for growth and the plants’ physiological responses at its current
phase of growth (Hurd et al., 1996). In macrophytes lacking roots, nutrient uptake
generally occurs across the blade surface, making surface area the measure that best
reflects the rate at which nutrients are taken up across the cell membrane (Hurd et al..

Literature Review

39

1996). Advection ensures a near continual supply and availability of nutrients to
macrophytes directly through the replenishment of water, and indirectly through the
supply to the sediments (Dawson, 1988).
Phosphorous is a nutrient that is mainly associated with organic matter and fine
mineral particles of aluminum, calcium, and iron compounds. As a result, coarsetextured sediments contain much lower concentrations of nitrogen and phosphorous than
fine-textured sediments (Sand-Jensen, 1998). Temporary sediment retention during
summer exerts a strong influence on sediment processes and reduces the transport of
nitrogen and phosphorous to downstream lakes and estuaries (Sand-Jensen, 1998).
Phosphorous and nitrogen are generally considered the most critical nutrients for
autotrophic production, although other elements (e.g., iron, potassium, and silica) can
limit growth under certain circumstances. In nutrient-poor freshwaters, phosphorous is
usually the principal nutrient limiting autotrophic production, whereas is in temperate
marine waters, nitrogen is the more common limiting nutrient (Chambers et al., 1999).
The relatively large and slow-growing macrophytes also have a much lower nitrogen and
phosphorus requirement per unit carbon (biomass) than the much smaller phytoplankton
(Kalff, 2002). When phosphorous is not limiting, temperature exerts the dominant
control over algal growth rates (Bothwell, 1988).
Grazing, Predation, and Competition
It is also possible for grazers and predators to have an effect on macrophyte
productivity. Although, direct grazing is generally not a controlling factor in the overall
distribution and biomass of freshwater plants (Chambers et al., 1999). The higher
structural carbon content appears to be responsible for making macrophytes as a group
less desirable to herbivores than phytoplankton or periphyton (Kalff, 2002).

Literature Review

40

Additionally, if resources are plentiful, there is no competition, regardless of the amount
of niche (not canopy) overlap between individual plants as well as different macrophyte
species (Abrams, 1980). Alternatively, if there is territoriality or if species actively alter
their resource utilization to avoid a competitor, competition may occur in spite of the fact
that there is little or no niche overlap (Abrams, 1980). Furthermore, it has been said that
abiotic processes dominate lotie processes when flows are highly energetic, whereas
biotic interactions such as competition or predation tend to occur in more predictable
flow environments (Poff and Ward, 1989). Competition can be defined as a contest
between plants for environmental resource, when that resource is limiting, and can be
identified by growth inhibitions in one species that are caused by another (Agami and
Waisel, 2002). For example, floating macrophytes (e.g., Azolla) can block light from
submerged species (e.g., Elodea) thus decreasing its productivity (Forchhammer, 1999).
Allelopathy is another form of competition whereby one organism produces a chemical
substance that inhibits the growth of another organism, giving the first species the
competitive advantage (Gopal and Goel, 1993). For example in Ceratophyllum
demersum emits a chemical which limit the growth of cyanobacteria (Gross et al., 2003).
Thus, competition is a complicated phenomena which is highly dependent on nutrient
acquisition; generally a function of flow.
Flow in Macrophyte Canopies
Effect of Macrophytes on Flow
Changes in water flow or velocity can alter the biomass and species composition
of submerged plant communities, however, aquatic plants can also alter flow patterns.
Marine and freshwater macrophytes can modify their physiochemical environment by

Literature Review

41

slowing water flow, trapping sediments, and altering temperature and water chemistry
profiles (Ackerman and Okubo, 1993; Chambers et al., 1999). These modifications occur
because the nature of flow is changed as it enters and passes through plant canopies.
Ackerman and Okubo (1993) have shown that upon entering the canopy, the velocity of
the ambient current is considerably reduced. They also demonstrated that flow within the
mid-portion of the canopy is consistent in terms of average velocity, with a slight
increase in speed below the region of maximum leaf area. Accordingly, turbulence in the
flow entering the canopy is suppressed by the consumption of energy by the
hydrodynamic drag forces of large, divided fronds (Anderson and Charters, 1982). The
drag is also reduced in canopies due to the protection of neighbouring plants (SandJensen, 2003). In short, because turbulence in the incoming flow is damped, the net
effect is a smooth internal and exiting flow at velocities below a critical value (6 - 1 2 cms'^
depending on the spacing of the branches), and an abrupt transition to turbulent flow at
velocities above (Anderson and Charters, 1982). Anderson and Charters (1982) also
demonstrated that the decrease in transition velocity from single plants to groups of
plants suggests that the transition depends on the density of branches in the flow path, not
only on a single dimension characteristic of the plant’s morphology. This is evident in
the modification of turbulence in the flow passing though an algae with bushy
morphology. If the flow entering the plant mass is turbulent, the turbulence will be
damped as the water flows through the plant. This occurs because the dense morphology
of the plant reduces the velocity within the plant structure. Conversely, turbulence is
generated in the flow within the plant canopy because small branches act a roughness
elements that trip the boundary layer creating an area of turbulence close to the plant

Literature Review

42

surface (Anderson and Charters, 1982). Additionally, it has been shown that different
macrophyte morphologies affect the flow within the canopy differently due to their
differences in stem flexibility (Shi and Hughes, 2002).
The plastic morphological response of macrophytes to water motion has prompted
the suggestion that by changing their morphology, submerged plants are able to modify
their hydrodynamic environment and thereby reduce drag or diffusional stresses (Hurd et
al., 1997). In a plant canopy, most of the flow enters through the upstream margin,
whereas the deflected flow moves above and along the sides of the plant canopy. The
high macrophyte density at the surface of the canopy develops in the unidirectional flow
of the shallow streams because the shoots and leaves bend over becoming very dense at
the canopy surface (Sand-Jensen and Pedersen, 1999). Sand-Jensen and Mebus (1996)
demonstrated that as water velocity is reduced within these macrophyte patches, the
velocity is accelerated around the patches to maintain the downstream discharge (i.e.,
continuity) of water, resulting in steeper vertical velocity gradients at the patch surface
than those encountered above the sediment surface upstream of the patch. Hurd and
Stevens (1997) showed that since small, finely branched plants slow down and dampen
turbulence, conditions within the plant body at low mainstream flows could be ideal for
the formation of thick diffusion boundary layers. Thus, for small, branched plants, the
potential for diffusion-limited productivity is much greater than that for larger, broadly
branched macrophytes.
Due to the fact that plants are flexible and interact with the flow, branching
patterns, leaf/stem length and stiffness will influence the effective plant surface area
exposed perpendicularly to the flow, and, thereby, the effective pressure drag. However,

Literature Review

43

due to the uncertainty of estimates of typical plant parameters, transition from turbulent
to laminar flow, and the flexibility of the plant canopies, it is very difficult to predict or
model turbulence since flow rates and boundary layer thickness change temporally
(Sand-Jensen and Pedersen, 1999). Sand-Jensen and Pedersen (1999) indicate that these
scaling parameters are important because the flow patterns generated or modified by the
presence of the plants have strong implications for: ( 1 ) plant metabolism, physical
resistance and canopy development; (2 ) sedimentation and resuspension of sediment
particles; and (3) the growth and survival of micro-organisms and invertebrates on plant
surfaces or on sediments shielded by the canopies.
Disturbance
Flow may be disrupted in streams due to unpredictable peak flows, which may
move the substratum, wash away individual organisms, and cause high mortality to biota
(Winterbottom et al., 1997). Hydraulic disturbance by sudden increases in flow may be
the major mechanism controlling differences in biomass and structure between streams.
The type and duration of disturbance may be more important to stream ecosystems than
factors such as nutrient levels or plant/animal interactions due to biomass removal (Biggs
and Thomsen, 1995). Extreme hydraulic forces accompanying spates (i.e., sudden
floods) can erode organisms from the stream bed, particularly where the substrate is
moved. Consequently, flow structures in the water column determine the likelihood that
entrained particles and benthic organisms will be transported out of the reach (Lancaster
and Hildrew, 1993). The degree of disturbance of the plant communities for a given peak
flow depends on the properties of both the plants and the habitat. At the plant level, the
physical structure of the organism (which determines the drag properties), the strength of

Literature Review

44

the root systems, intercellular connections, and how flexible the organism is, are
important factors determining the impact of disturbance. At the habitat level, the age of
the community and the shear stress to which it is already acclimatized contribute to the
biotic resistance of the community (Biggs, 1996). Along these lines, the degree of habitat
resistance is primarily a function of bed sediment stability or, more specifically, the
ability of given sediments to resist being carried along in the current (Biggs, 1996).
Velocity Profiles
The complex spatial distribution of roughness elements among plant stands,
stream bed and banks, and the differences in flow velocity outside and inside plant
canopies, may result in variable velocity profiles and the distortion of logarithmic profiles
(Sand-Jensen and Pedersen, 1999). This variability is representative of higher slopes and
higher hydraulic roughness leading to a more turbulent flow for the same mean velocity.
The vertical hydraulic gradients and deviations from the mean current flow created by
streambed features such as riffles and bars may play a strong role in the distribution and
nutrition of lotie macrophytes (White and Hendricks, 2000). Another factor influencing
velocity profiles is seasonal variability in the rate of discharge, however, the seasonal
growth of plants can in turn modify the speed or velocity of this flow. Major changes in
flow may eliminate or strongly suppress the growth of a plant population. Plants adjust
to increased flow conditions by shedding their above-substratum parts or by rapid
reinvasion from less vulnerable areas through the production of numerous propogules or
seeds (Dawson, 1988).

Literature Review

45

Sediments
The creation of a low energy environment above the sediments within macrophyte
patches leads to the retention of fine mineral and organic particles (Petticrew and Kalff,
1992; Sand-Jensen, 1998). This trapping of particles by macrophytes also produces
sediments with a higher organic content and smaller grain size within the bed than in
nearby uncolonized areas (Chambers et al., 1999). Increased current velocity, however,
results in river beds with larger sediment particles that are usually less nutrient rich, and
less densely packed than river beds that experience slower current velocity. For this
reason, aquatic macrophytes tend to grow better in low flows, either because of increased
sediment nutrient concentrations or due to the increased sediment stability (Chambers et
al., 1991). Consequently, both macrophyte diversity and the distribution of macrophytes
is positively correlated to the substratum type (Baattrup-Pedersen and Riis, 1999)
demonstrating the importance of having fine sediments settle out of the water column
(French and Chambers, 1996).
Baattrup-Pedersen and Riis (1999) showed that emergent macrophyte species are
primarily assoeiated with finer-textured substrata (e.g. mud, peat and fine sand), whereas
submerged macrophyte species are primarily associated with the coarser-textured
substrata (e.g. coarse sand, stone, and gravel). For macrophytes, substrate type and
coheSiveness also affect the strength of below-ground anchorage. In sediments that are
not yet compact, the long-rooting plants (e.g. Sparganium erectum) are best adapted to
resist spates. On coarser gravels the plants usually have a dense interweaving of shallow,
often horizontal, roots (e.g. Rannunculus spp.), which curl around particles. These

Literature Review

46

different methods of anchoring can stabilize the habitat and increase the overall resistance
of a site to flow changes (Biggs, 1996).
Methodologies
Given the importance of fluid dynamics to aquatic plants, it is important to
characterize and measure the fluid dynamic conditions in and around macrophyte
canopies. A variety of flow visualization techniques exist that allow the observer to see
the flow patterns, and depending on the complexity of the technique, measurement
velocity and turbulence. Additionally, there are several methods of measuring velocity
from which other quantities such as shear stress can be derived. It is important to have an
accurate representation of flow conditions since they are one of the main factors
regulating freshwater macrophyte productivity.
Flow Visualization
One of the most effective ways of assessing fluid flow is by observing streaklines
of dye. Dye is injected into the flow upstream of the working section through an Lshaped hypodermic tube. The tube has to be so thin that its wake is smooth and laminar
(Anderson and Charters, 1982). This method was used by Sand-Jensen and Pedersen
(1999), whereby dye was injected into the flow upstream and along the sides of
macrophyte stands, as well as in the interior, and the movements of the dye were
observed to aid in the interpretation of the flow patterns. Nepf (1999) used a similar dye
plume technique to measure diffusivity in the field within an emergent stand of Spartina
altemiflora. When using dye tracking as a method of quantifying flow patterns, it is
important to record the water depth, canopy height, percentage of the water column
occupied by vegetation, vegetation density, size and patchiness of the plant bed, wind

Literature Review

47

intensity and direction, and waves if possible and applicable (Koch and Verduin, 2001).
The disadvantage of flow visualization with dye is it only allows for an estimation of
average flows over relatively broad spatial scales when used in the field. Thus, when
smaller scale questions are asked, more sophisticated techniques (such as PIV as is
discussed in a later section) need to be employed (e.g., Koch and Verduin, 2001).
Alternatively, according to Hurd et al. (1997) another way to visualize the flow is
by using time-lapse photography of two tracers, whereby a thin mixture of reflective resin
is added to the tank and used to create particle streaks. Using this technique, Hurd et al.
(1997) were able to see that the gradient region of the velocity profile becomes thinner
with increasing free-stream velocity. It is possible to compare these observations with
predicted values from the flat-plate boundary layer theory (Hurd et al., 1997). Neutrally
buoyant particles can also be added to the water, and be illuminated by a laser light sheet.
Images of the illuminated particles are taken using a high-resolution video camera
equipped with a macro lens, mounted perpendicular to the light sheet (Stamhuis and
Yideler, 1998). Mica Pearlescence flakes acting as tracers are another way to assess flow
patterns. These particles are generally too fine to be used in velocity analysis but have
the advantage that their plate-like nature aligns with velocity gradients so that regions of
velocity variation, including turbulence, are visually represented (Hurd and Stevens,
1997). For example, in steep velocity gradients, the mica particles minimize drag by
aligning with the shear. There is little reflection of light off of the aligned particles, and
the viewer sees a dark patch of fluid; the black regions in the photographs therefore
indicate shear. In regions where velocity gradients are weak, the mica particles are

Literature Review

48

aligned randomly and scatter light in all directions, appearing silvery white (Hurd and
Stevens, 1997).
Velocimetry
Particle-image velocimetry (PIV) is a powerful method to measure turbulence as
well as coherent vortices in shallow water flows. This method offers spatial information
related to the velocity field (Nezu and Onitsuka, 2001). A laser beam is expanded with a
cylindrical lens system to form the plane light sheet whose thickness can be made less
than 1 mm. Tracer particles are illuminated only when moving through this light sheet,
and their images are recorded by a camera whose direction of observation is usually
facing the sheet (Khalili et al., 2001). In the most frequently used PIV technique, two
short exposures are taken, separated by a short time interval. The velocity measurement
is thereby reduced to a recording of the displacement of individual particles during that
time interval (Khalili et al., 2001).
An acoustic Doppler velicometer (ADV) can be used to measure velocity,
turbulence, Reynolds stress and turbulence intensity for a given volume adjacent to or
above the substratum (Bouckaert and Davis, 1998), as well as around macrophyte blades
(Rybicki et al., 1997; reviewed in Hurd, 2000). The probe emits a series of acoustic
pulses and records the reflection of these pulses from particles suspended in water
(Khalili et al., 2001). The acoustic sensor consists of a transmitting transducer
surrounded by two (2D) or three (3D) receiving transducers, usually mounted on short
arms. The transmitted beams are typically orientated so that the sampling volume is
located below the probe. This volume has a cylindrical shape, with a diameter of
approximately 0.5 cm and a height of about 1 cm (Khalili et al., 2001).

Literature Review

49

Velocity can also be measured with cylindrical hot-wire anemometers, to help
with the interpretations of flow visualization around and within macrophyte stands (SandJensen and Pedersen, 1999). The working principle of the thermal anemometer is based
on heat transfer from an immersed, electrically heated sensor to a fluid. In homogeneous
fluids with constant temperature and pressure, the velocity of the fluid passing the sensor
determines the rate of heat loss from the heated element; that is, more heat is released to
the fluid at higher velocities (Khalili et al., 2001).
Approximations and Derived Measurements
To determine values for mean shear stress, velocity profiles are measured, from
which velocity is regressed against the logged depth values from the log-linear part of the
velocity profile (Biggs and Thomsen, 1995; reviewed in Ackerman and Hoover, 2001).
However, non-logarithmic layers may occur in strongly accelerating or decelerating flow
like those found in river reaches which narrow or widen rapidly. The logarithmic profile
may also be distorted by spatial variation in the mixture of roughness types on the bed, or
by the extreme bed roughness that occur over dunes and large rocks (Carling, 1992;
Hoover and Ackerman, in Press). In these circumstances it is often possible to obtain
near log-normal distributions of velocity from close to the bed, which can be used to
estimate the local bed shear stress (Carling, 1992). The velocity of the fluid is zero at the
surface of any solid object and a velocity gradient (du/dz) must form perpendicular to the
solid surface. This gradient (as extrapolated from the velocity profile) is the change in
velocity with the change in distance perpendicular to the solid surface. The shear stress in
laminar flow is thus expressed as: T = p (du/dz); where p is the viscosity (Vogel, 1994;
Ackerman, 1997; reviewed in Hurd, 2000). However, things are complicated in turbulent

Literature Review

50

flow because flow moves in both the horizontal and the vertical direction. To account for
this the equation becomes f =

where f = l ^ —

is the kinematic eddy

viscosity (Kundu, 1990). This is simplified into the law of the wall w = — In — ,
^

V^ o y

where m* is the friction velocity and r - p u l .
A Preston-static tube can be used to measure shear stress in laboratory as well as
field settings (Ackerman and Hoover, 2001). This technique involves the use of a surface
mounted Pitot tube to measure the total pressure at the boundary and a static pressure tap.
The Preston tube must be small enough to be within the wall layer, pressure differences
must be large, and the static pressure must be constant and measured close to the dynamic
pressure tap (Ackerman and Hoover, 2001). This technique is however limited in that the
flow needs to be as unidirectional as possible, and the water needs to be fast flowing
since this method depends on measuring the dynamic pressure of the flow (Ackerman
and Hoover, 2001).
Other Measurements
Drag, according to Schutten and Davy (2000), can be measured by attaching each
individual shoot to a spring balance using a clamp of neutral buoyancy, with known and
very small resistance, before submerging the blade into the flume current. The resulting
tension is then read from the spring balance and corrected for the resistance arising from
the clamp. Forces measured with the spring balance could potentially have a vertical
component included because of the angle between the current and the plane of
measurement; however, any such component would have been transferred to the tension
only as the sine of that small angle.

Literature Review

51

The dispersion of particles released and captured in a macrophyte canopy were
used to measure the eddy diffusivity (Ackerman, 2002). This method can also be used
with plastic models of plants as the capturing surface in order to examine the eddy
diffusivity around different shapes, and at different heights in the water column (Harvey
et al., 1995). A similar technique is to inject microparticles over a set period of time,
after which those attached to glass rods (covered in vacuum grease) are counted under a
black light. Once again, this is a technique used to determine the particle distribution and
thus, the eddy viscosity, across the flume section (Harvey et al., 1995).
A technique that provides a relative estimate of water motion between and within
sites is to measure the dissolution of a substance such as plaster, sugar, or benzoic acid
from a module attached to the macrophyte surface (Angradi and Hood, 1998; Koehl and
Alberte, 1988; reviewed in Hurd, 2000). Rough estimates of average shear velocities
along surfaces of blades of different morphologies exposed to the same field conditions
were made by measuring the weight loss during a 10 min interval of candy (e.g.. Life
Savers) sewn to blades (Koehl and Alberte, 1988). This method requires that a
calibration model be developed in a flume to relate mass loss of plaster, sugar or benzoic
acid standards to water velocity and temperature. This model can then be used to
calculate water velocity based on in situ loss of these substance standards (Angradi and
Hood, 1998). This method is, however, very limited because the flow variables
addressed using dissolution are diverse and are often nonspecific. For example, one
cannot discern between the measurement of current velocity, flow intensity, turbulence
intensity, or water motion (Porter et al., 2000).

Literature Review

52

Conclusion
The vegetation of aquatic systems is defined by the movement of water, which
influences plant form, dominates the growth-controlling factors (e.g. nutrients and light),
and defines the canopy structure. Thus, different flows will directly determine the
presence (or absence), and location of instream vegetation (Dawson, 1988). On a smaller
scale, it can be seen that water motion affects an aquatic plant through the transport of
mass, momentum, and energy from the flowing water through the boundary layer to the
plant’s surface. The transport rates depend strongly on whether the flow in the boundary
layer is laminar or turbulent. This is because the transport rates in turbulent flow with
eddy diffusion are much larger than those in laminar flow with molecular diffusion
(Anderson and Charters, 1982). Consequently, it is important for macrophytes to be able
to alter morphologically to flow conditions because, different morphologies (i.e., bladed,
whorled, and dissected leaves) potentially have different implications on the flow. This
is relevant since it has been documented that there is a close relationship between
environment and leaf structure and function (Rascio, 2002). Additionally, since plant
form is important in creating and defining habitat in aquatic systems, it is important to
examine the fluid dynamics surrounding macrophytes at a small scale using flow
visualization techniques in order to determine what is happening around individual
macrophytes to make implications to an ecological scale. The ecological scale provides
an integrated response to a broad range of disturbances (Bunn et al., 1999). Intact
ecosystems are required as references to compare the effectiveness of restoration
programs, and as sources of natural genetic material for local stream reaches in need of
restoration (Kauffman et al., 1997). Unfortunately, undisturbed ecosystems are becoming

Literature Review
rare, and for many rivers it is due to anthropogenic processes (e.g. timber harvest,
livestock grazing, agriculture, and urbanization; Poff et al., 1997) which alter flow
regimes.

53

Literature Review

54

Literature Cited
Abrams, P. 1980. Some comments on niche overlap. Ecology 61(l):44-49.
Ackerman, J.D. 1995. Convergence of filiform pollen morphologies in seagrasses:
functional mechanisms. Evolutionary Ecology. 9:139-153.
Ackerman, J.D. 1997. Submarine pollination in the marine angiosperm Zostera marina
(Zosteraceae). I. The influence of floral morphology on fluid flow. Journal o f Botany
84(8).T 099-1109.
Ackerman, J.D. 2000. Abiotic pollen and pollination: ecological, functional, and
evolutionary perspectives. Plant Systematics and Evolution 222:167-185.
Ackerman, J.D. 2002. Diffusivity in a marine macrophyte canopy: Implications for
submarine pollination and dispersal. American Journal o f Botany 89(7): 1119-1127.
Ackerman, J.D. and T.M. Hoover. 2001. Measurement of local bed shear stress in
streams using a Preson-static tube. Limnology and Oceanography 46(8):2080-2087.
Ackerman, J.D., and A. Okubo. 1993. Reduced mixing in a marine macrophyte canopy.
Functional Ecology 7:305-309.
Agami, M., and Y. Waisel. 2002. Competitive relationships between two water plant
species: Najas marina L. and Myriophyllum spicatum L. Hydrobiologia. 482:197-200.
Anderson, S.M., and A C. Charters. 1982. A fluid dynamics study of seawater flow
though Gelidium nudifrons. Limnology and Oceanography 27(3):399-412.
Angradi, T., and R. Hood. 1998. An application of the plaster dissolution method for
quantifying water velocity in the shallow hyporheic zone of an Appalachian stream
system. Freshwater Biology 39:301-315.
Baattrup-Pedersen, A. and T. Riis. 1999. Macrophyte diversity and composition in
relation to substratum characteristics in regulated and unregulated Danish streams.
Freshwater Biology 42:375-385.
Barrat-Segretain, M-H. 2001. Biomass allocation in three macrophyte species in relation
to the disturbance level of their habitat. Freshwater Biology 46:935-945.
Biggs, B.J.F. 1996. Hydraulic habitat of plants in streams. Regulated Rivers: Research
and Management 12:131-144.
Biggs, B.J.F., and H.A. Thomsen. 1995. Disturbance of stream periphyton by
perturbations in shear stress: time to structural failure and differences in community
resistance. Journal o f Phycology 31:233-241.

Literature Review

55

Borchardt, M.A. 1994. Effects of flowing water on nitrogen- and phosphorous-limited
photosynthesis and optimum N:P ratios by Spirogyra Fluviatilis (Charophyceae).
Journal o f Phycology 30:418-430.
Bothwell, M.L. 1988. Growth rate responses of lotie periphytic diatoms to experimental
phosphorous enrichment: The influence of temperature and light. Canadian Journal o f
Fisheries and Aquatic Sciences 45:261-270.
Bouckaert, F.W. and J. Davis. 1998. Microflow regimes and the distribution of
macroinvertebrates around stream boulders. Freshwater Biology 40:77-86.
Bunn, S.E., P.M. Davis, and T.D. Mosisch. 1999. Ecosystems measures of river health
and their response to riparian and catchment degradation. Freshwater Biology 41:333345.
Carling, P.A. 1992. The nature of the fluid boundary layer and the selection of parameters
for benthic ecology. Freshwater Biology 28:273-284.
Chambers, P.A., R E. DeWreede, E.A. Irlandi, and H. Vandermeulen. 1999. Management
issues in aquatic macrophyte ecology: a Canadian perspective. Canadian Journal o f
Botany 77:471-487.
Chambers, P.A., E.E. Prepas, H R. Hamilton, and M.L. Bothwell. 1991. Current velocity
and its effect on aquatic macrophytes in flowing waters. Ecological Applications
l(3):249-257.
Cheer, A.Y.L., and M.A.R. Koehl. 1987. Paddles and rakes: Fluid flow through bristled
appendages of small organisms. Journal o f Theoretical Biology 129:17-39.
Christian, D., and Y.P. Sheng. 2003. Relative influence of various water quality
parameters on light attenuation in Indian River Lagoon. Estuarine, Coastal and Shelf
Science. 57:961-971.
Coutanceau, M., and J. Defaye. 1991. Circular cylinder wake configurations: A flow
visulization survey. Applied Mechanics Reviews. 44(6):255-305.
Creed, J.C., T.A. Norton, and J.M. Kain. 1996. Are neighbours harmful of helpful in
Eucus vesicolosus populations? Marine Ecology Progress Series 133:191-201.
Cronin, G., and D.M. Lodge. 2003. Effects of light and nutrient availability on the
growth, allocation, carbon/nitrogen balance, phenolic chemistry, was resistance to
herbivory to two freshwater macrophytes. Oecologia 137:32-41.
Crossley, M.N., W.C. Dennison, R.R. Williams, and A.H. Wearing. 2002. The interaction
of water flow and nutrients on aquatic plant growth. Hydrobiologia 489:63-70.

Literature Review

56

Davis, J.A., and L.A. Barmuta. 1989. An ecologically useful classification of mean and
near-bed flows in streams and rivers. Freshwater Biology 21:271-282.
Davison, I.R. 1991. Environmental effects on algal photosynthesis: temperature. Journal
o f Phycology 27:2-8.
Dawes, C.J. 1981. Marine Botany. John Wiley and Sons, Inc. U.S.A. 628 pp.
Dawson, F.H. 1988. J.J. Symoens (ed.). Vegetation o f Inland Waters. Kluwer Academic
Publishers, Dordrecht. Netherlands. Pp. 283-309.
Den Hartog, C. 1982. Architecture of macrophyte-dominated aquatic communities.
Studies on Aquatic Vascular Plants, ed. Symoens, J.J., S.S. Hooper and P. Compere.
Royal Botanical Society of Belgium, Brussels. Pp. 222-234.
Dennison, W.C. 1987. Effects of light on seagrass photosynthesis, growth and depth
distribution. Aquatic Botany 27:15-26.
Denny, M.W. 1993. Air and Water: The Biology and Physics o f Life’s Media. Princeton
University Press, U.S.A. 341 pp.
Denny, M.W., B.P Gaylord, and E.A. Cowen. 1997. Flow and flexibility II. The roles of
size and shape in determining wave forces on the bull kelp Nereocystis luetkeana.
Journal o f Experimental Biology 200:3165-3183.
Denny, M., B. Gaylord, B. Helmuth, and T. Daniel. 1998. The menace of momentum:
dynamic forces on flexible orgamisms. Limnology and Oceanography 43(5):955-968.
Denny, M., and L. Robertson. 2002. Blade motion and nutrient flux to the kelp, Eisenia
arborea. Biological Bulletin 203:1-13.
Ennos, A.R. 1999. The aerodynamics and hydrodynamics of plants. Journal o f
Experimental Biology 202:3281-3284.
Eugster, W., and R. Hesterberg. 1996. Transfer resistances of N 02 determined form eddy
correlation flux measurements over a litter meadow at a rural side on the Swiss
Plateau. Atmospheric Environment. 30(8): 1247-1254.
Falkowski, P., and J.A. Raven. 1997. Aquatic Photosynthesis. Blackwell, Oxford, 375pp.
French, T.D., and P.A. Chambers, 1996. Habitat partitioning in riverine macrophyte
communities. Freshwater Biology 36:509-520.
Forchhammer, N.C. 1999. Production potential of aquatic plants in systems mixing
floating and submerged macrophytes. Freshwater Biology. 41:183-191.

Literature Review

57

Gerard, V.A. 1987. Hydrodynamic streamlining of Laminaria saccharina Lamour. in
response to mechanical stress. Journal o f Experimental Marine Biology and Ecology
107:237-244.
Gerard, V.A., and K.H. Mann. 1979. Growth and production of Laminaria longicruris
(Phaeophyta) populations exposed to different intensities of water movement. Journal
o f Phycology 15:33-41.
Gopal, B., and U. Goel. 1993. Competition and allelopathy in aquatic plant communities.
The Botanical Review 59(3): 155-210.
Gregg, W.W., and F.L. Rose. 1983. The effects of aquatic macrophytes on the stream
micro-environment. Aquatic Botany 14:309-324.
Gross, E.M., D. Erhard, and E. Ivanyi. 2003. Allelopathic activity of Ceratophyllum
demersum L. and Najas marina ssp. intermedia (Wolfgang) Casper. Hydrobiologia.
506-509:583-589.
Gutschick, V.P. 1999. Research reviews: biotic and abiotic consequences of differences
in leaf structure. New Phytologist 143(1):3-18.
Hart, D.D., B.D. Clark, and A. Jasentuliyana. 1996. Fine-scale field measurement of
benthic flow environments inhabited by stream invertebrates. Limnology and
Oceanography 41(2):297-308.
Harvey, M., E. Bourget, and R.G. Ingram. 1995. Experimental evidence of passive
accumulation of marine bivalve larvae on filamentous epibenthic structures.
Limnology and Oceanography 40(1):94-104).
Hinze, J.O. 1987. Turbulence 2"^ ed. McGraw-Hill, Inc.
Hoover, T.M., and J.D. Ackerman. In Press. Near-bed hydrodynamics measurements
above holders in shallow torrential streams: Implications for stream biota. Journal o f
Environmental Engineering and Science.
Hurd, C.L. 2000. Water motion, marine macroalgal physiology, and production. Journal
o f Phycology 36:453-472.
Hurd, C.L., P.J. Harrison, and L.D. Druehl. 1996. Effect of seawater velocity on
inorganic nitrogen uptake by morphologically distinct forms of Macrocystic
integrifolia from wave-sheltered and exposed sites. Marine Biology 126:205-214.

Literature Review

58

Hurd, C.L., C.L. Stevens, E.L. Bernard, G.A. Lawrence, and P.J. Harrison. 1997.
Visualization of seawater flow around morphologically distinct forms of the giant kelp
Macrocystis integrifolia from wave-sheltered and exposed sites. Limnology and
Oceanography 42(1): 156-163.
Hurd, C.L., and C.L. Stevens. 1997. Flow visualization around single- and multiplebladed seaweeds with various morphologies. Journal o f Phycology 33:360-367.
Idestam-Almquist, J., and L. Kautsky. 1995. Plastic responses in morphology of
Potamogeton pectinatus L. to sediment and above-sediment conditions at two sites in
the northern Baltic proper. Aquatic Botany 52:205-216.
James, W.F., E.P. Best, and J.W. Barko. 2004. Sediment resuspension and light
attenuation in Peoria Lake: can macrophytes improve water quality in this shallow
system? Hydrobiologia. 515:193-201.
Kalff, J. 2002. Limnology: Inland Water Ecosystems. Prentice Hall, Upper Saddle River,
New Jersey.
Kauffman, J.B., R.L. Beschta, N. Otting, and D. Lytjen. 1997. An ecological perspective
of riparian and stream restoration in the western United States, Watershed Restoration
22(5): 12-24.
Khalili, A., M. Huettel, and W. Merzkirch. 2001. Fine-scale flow measurements in the
benthic boundary layer. Pp. 44-77 In The Benthic Boundary Layer Transport
Processes and Biogeochemistry, ed. Boudreau, B.P. and B.B. Jorgensen. Oxford
University Press.
Kirk, J.T.O. 1994. Light and Photosynthesis in Aquatic Systems. Cambridge University
Press, Cambridge. 509 pp.
Koch, F.W. 1993. The effect of water flow on photosynthetic processes of the alga Ulva
lactuca L. Hydrobiologia 260/261:457-462.
Koch, F.W. 1994. Hydrodynamics, diffusion-boundary layers and photosynthesis of the
seagrasses Thalassia testudinum and Cymodocea nodosa. Marine Biology 118:767776.
Koch, F.W. and J.J. Verduin. 2001. Measurements of physical parameters in seagrasss
habitats. Pp. 325-344 In Global Seagrass Research Methods, ed. Short, F T., R.G.
Coles, and C.A. Short. Elsevier Science B.V.
Koehl, M.A.R. 1996. When does morphology matter? Annual Review o f Ecology and
Systematics 27:501-542.

Literature Review

59

Koehl, M.A.R., and R.S. Alberte. 1988. Flow, flapping, and photosynthesis of
Nereocystis luetkeana: a functional comparison of undulate and flat blade
morphologies. Marine Biology 99:435-444.
Kundu, P.K. 1990. Fluid Mechanics. Academic Press, Inc. pp. 638.
Lancaster, J., and A.G. Hildrew. 1993. Characterizing in-stream flow réfugia. Canadian
Journal o f Fisheries and Aquatic Sciences 50:1663-1675.
Lieffers, V.J., C. Messier, K.J. Stadt, F. Gendron and P.G. Comeau. 1999. Predicting and
managing light in the understory of boreal forests. Canadian Journal o f Forest
Research 29:796-811.
Leonard, G.H., J.M. Levine, P R. Schmidt, and M.D. Bertness. 1998. Flow-driven
variation in intertidal community structure in a Maine estuary. Ecology 79:1395-1411.
Losee, R.F., and R.G. Wetzel. 1993. Littoral flow rates within and around submersed
macrophyte communities. Freshwater Biology 29:1-11.
Mathieu, J., and J. Scott. 2000. An Introduction to Turbulent Flow. Cambridge
University Press.
Nepf, H.M. 1999. Drag, turbulence, and diffusion in flow through emergent vegetation.
Water Resources Research 35(2):479-489.
Nezu, I., and K. Onitsuka. 2001. Turbulent structures in partly vegetated open-channel
flows with LDA and PIV measurements. Journal o f Hydraulic Research 39(6):629642.
Niinemets, U., and S. Fleck. 2002. Petiole mechanics, leaf inclination, morphology, and
investment in support in relation to light availability in the canopy of Liriodendron
tulipifera. Oecologia 132:21-33.
Niklas, K.J. 1992. Plant Biomechanics: An Engineering Approach to Plant Form and
Function. The University of Chicago Press, Chicago. 607 pp.
Niklas, K.J. 1997. The Evolutionary Biology o f Plants. The University of Chicago Press,
Chicago. 449 pp.
Nowell, A.R.M., and P.A. Jumars. 1984. Flow environments of aquatic benthos. Annual
Review o f Ecology and Systematics 15:303-328.
Petticrew, E.L., and J. Kalff. 1992. Water flow and clay retention in submerged
macrophyte bed. Canadian Journal o f Fisheries and Aquatic Science. 49(12):24832489.

Literature Review

60

Philbrick, C.T., and D.H. Les. 1996. Evolution of aquatic angiosperm reproductive
systems. Bioscience. 46(Il):813-826.
Poff, N.L., and J.V. Ward. 1989. Implications of streamflow variability and predictability
for lotie community structure: a regional analysis of streamflow patterns. Canadian
Journal o f Fisheries and Aquatic Sciences 46:1805-1817.
Poff, N.L., J.D. Allan, M.B. Bain, J.R. Karr, K.L. Prestegaard, B.D. Richter, R.E. Sparks,
and J.C. Stromberg. 1997. The natural flow regime a paradigm for river conservation
and restoration. Bioscience 47(ll):769-782.
Porter, E.T., L.P. Sanford, and S.E. Suttles. 2000. Gypsum dissolution is not a universal
integrator of ‘water motion’. Limnology and Oceanography 45(1):145-158.
Rascio, N. 2002. The underwater life of secondarily aquatic plants: some problems and
solutions. Critical Reviews in Plant Science 21(4):401-427.
Raven, P.H., R.E. Evert, and S.E. Eichhom. 1999. Biology o f Plants. 6* ed. W.H.
Freeman and Company, New York. 944 Pp.
Reimer, D.N. 1993. Introduction to Freshwater Vegetation. Krieger publishing company,
Malabar, Florida. Pp. 205.
Rybicki, N.B., H.L. Jenter, V. Carter, R.A. Baltzer, and M. Turtora. 1997. Observations
of tidal flux between a submersed aquatic plant stand and the adjacent channel in the
Potomac River near Washington, D C. Limnology and Oceanography 42:307-317.
Sand-Jensen, K. 1998. Influence of submerged macrophytes on sediment composition
and near-bed flow in lowland streams. Freshwater Biology 39:663-679.
Sand-Jensen, K. 2003. Drag and reconfiguration of freshwater macrophytes. Freshwater
Biology 48:271-283.
Sand-Jensen, K., and J.R. Mebus. 1996. Fine-scale patterns of water velocity within
macrophyte patches in streams. Oikos 76:169-180.
Sand-Jensen, K., and O. Pedersen. 1999. Velocity gradients and turbulence around
macrophyte stands in streams. Freshwater Biology 42:315-328.
Santamaria, L. 2002. Why are most aquatic plants widely distributed? Dispersal, clonal
growth and small-scale heterogeneity in a stressful environment. Acta Oecologica
23:137-154.
Schutten, J., and A.J. Davy. 2000. Predicting the hydraulic forces on submerged
macrophytes from current velocity, biomass and morphology. Oecologia 123:445-452.

Literature Review

61

Sculthorpe, C.D. 1967. The Biology o f Aquatic Vascular Plants. Edward Arnold, London
(Reprinted 1985 Koeltz Scientific, Konigstein), 610 pp.
Sheath, R.G., and J.A. Hambrook. 1988. Mechanical adaptations to flow in freshwater
red algae. Journal o f Phycology 24:107-111.
Shi, Z., and J.M.R. Hughes. 2002. Laboratory flume studies of microflow environments
of aquatic plants. Hydrological Processes 16:3279-3289.
Speck, O. 2003. Field measurements of wind speed and reconfiguration in Arundo donax
(Poaceae) with estimates of drag force. American Journal o f Botany 90(8): 1253-1256.
Stamhuis, E.J., and J.J. Videler. 1998. Burrow ventilation in the tube-dwelling shrimp
Callianassa subterranean (Decopoda: Thalassinidea). Journal o f Experimental
B/o/ogy 201:2159-2170.
Stevens, C.L., and C.L. Hurd. 1997. Boundary-layers around bladed aquatic
macrophytes. Hydrobiologia 346:119-128.
Stewart, H.L., and R.C. Carpenter. 2003. The effects of morphology and water flow on
photosynthesis of marine macroalgae. Ecology (84)11:2999-3012.
Strand, J.A., and S.E.B. Weisner. 2001. Morphological plastic responses to water depth
and wave exposure in and aquatic plant {Myriophyllum spicatum). Journal o f Ecology
89:166-175.
Thompson, P.A., and S.M. Troian. 1997. A general boundary condition for liquid flow at
solid surfaces. Nature 389:369-362.
Vesala, T. 1998. On the concept of leaf boundary layer resistance for forced convection.
Journal o f Theoretical Biology. 194:91-100.
Vogel, S. 1987. Flow-assisted mantle cavity refilling in jetting squid. Biological Bulletin
172:61-68.
Vogel, S. 1988. Life’s Devices The Physical World o f Animals and Plants. Princeton
University Press. Princeton, N.J.
Vogel, S. 1994. Life in Moving Fluids The Physical Biology o f Flow. 2"^ ed. Princeton
University Press, Princeton, N.J.
Watson, L., and M.J. Dalhoitz. 1992 (onwards). The families of flowering plants:
descriptions, illustrations, identification, and information retrival. Version: 6th
August 2003. http://delta-intkey.com.

Literature Review

62

Wells, C.L., and M. Pigliucci. 2000. Adaptive phenotypic plasticity: the case of
heterophylly in aquatic plants. Perspectives in Plant Ecology, Evolution and
Systematics 3(1):1-18.
Wheeler, W.N. 1988. Algal productivity and hydrodynamics - a synthesis. Progress in
Phycological Research 6:23-58.
White, D.S., and S.P. Hendricks. 2000. J.B. Jones and P.J. Mulholland (eds.) Lotie
macrophytes and surface-subsurface exchange processes Pp. 363-379 In Streams and
Ground Waters, ed. Jones, J.B., and P.J. Mulholland. Academic Press.
Wilson, C.A.M.E., T. Stoesser, P.D. Bates, and A. Batemann Pinzen. 2003. Open channel
flow through different forms of submerged flexible vegetation. Journal o f Hydraulic
Engineering 129(ll):847-853.
Winterbottom, J.H., S.E. Orton, A.G. Hildrew, and J. Lancaster. 1997. Field experiments
on flow réfugia in stream. Freshwater Biology 37:569-580.
Wright, J.P., W.S.C. Gurney, and C.G. Jones. 2004. Patch dynamics in a landscape
modifies by ecosystem engineers. Oikos 105:336-348.
Zhang, Y., and B. Malmqvist. 1997. Phenotypic plasticity in a suspension-feeding insect,
Simulium lundstromi (Diptera: Simuliidae), in response to current velocity. Oikos
78:503-510.

63

Dye Streakline

a)

Laminar

b)

Transitional

c)

Turbulent

Figure 2.1: a) Laminar flow as seen by the smooth dye streakline, b) transitional flow as
seen by the increasing waviness of the streakline, and c) turbulent flow as seen by the
choatic nature of the streakline.

64

a)

b)

c)

d)

e)

f)

Figure 2.2: The transition to turbulent flow of a fluid as it passes a circular cylinder seen
in cross section at different Reynolds numbers: (a) 1, (b) 10, (c) 13, (d) 26, (e)105, and (f)
150.
(From: http://nmm.media.mit.edu/student/95/aries/mas864/obstacles.html)

65

Figure 2.3: A rectangular block of fluid is deformed by shear stress, so two faces become
parallelograms (after Vogel, 1994).

66

Benthic Boundaiy Layer

Outer Layer
- dcxrdnated by transpot processes

a

I
I
8

I

Inertial Layer

Q

- dcmnated by turbulait stresses

Viscous Sublayer
- doirinated by viscous stresses

Figure 2.4: The benthic boundary layer concept. A velocity profile is presented from the
surface (U = 0) to the free stream (U = U«=; after Ackerman, 1989).

67

Figure 2.5: The four basic habitats of freshwater macrophytes: (A, B) free-floating; (C)
submerged; (D, F) floating-leaved; and (E, F) emergent (Wells and Pigliucci, 2000).

68

PODOSTEMOMACEAS.

Ttbâelii

Toi.

flfot ÉBIWb«. P ficnnr. CTimaf W Uft^m

1. FWawm&l,

Figure 2.6: Examples of caulescent macrophytes a) Elodea canadensis and b)
Myriophyllum spicatum (from; http://plants.ifas.ufl.edu/myrspi8.jpg; Vie Ramey,
University of Florida) ; rosette macrophytes c) Isoetes spp. (from:
http://www.anbg.gov.au/images/photo_cd/9J18G146043/006; Murray Fag, Australian
National Botanic Gardens) and d) Vallisneria americana', and a thalloid macrophyte e)
Tristicha alternifolia (From: http://biodiversity.uno.edu/delta/angio/www/podostem.htm;
Watson and Dalhoitz, 1992 onwards).

69

a)

b)

d)

Figure 2.1 \ Different arrangements of leaves around plant stems: a) opposite, b) alternate,
c) whorled, and d) rosulate. (From:
http://extension.oregonstate.edu/mg/botany/leaves3.html; 2004 Oregon Sate University)

70

Linear
Lanceolate
Oblong
///
\x

Elliptic
Ovate

/

Cordate

Figure 2.8: Morphologies of entire leaves ranging from linear to sagittate (cordate).
(From: http://www.csdl.tamu.edu/FLORAAVilson/tfp/veg/shapes3.gif)

71

Figure 2.9: An example of a petiole-like freshwater macrophyte, Nuphar advena (from:
http://plants.ifas.ufl.edu/nupadv7.jpg; Vic Ramey, University of Florida)

72

Leaf Margjln;

serrate

undulate

Figure 2.10: Different leaf margins that occur within entire leaf morphologies (From:
httD.7/collections.ic.gc.ca/gardens/Horticulture/The%20Structure%20of%20Plants.htm).

73

Figure 2.11: Fenestration occurs when small areas of interstitial tissue are lacking at
maturity as demonstrated in this leaf of Aponogeton madagascariensis (Photo by Patrick
Ragaz).

74

Figure 2.12: Dissected leaves (e.g. Ceratophyllum) occur when the leaf is subdivided.

75

1 cm

Figure 2.13; From left to right, the leaves represent the transition from submerged to
aerial leaf morphology observed in the vertically ascending stems of Proserpinaca
palustris. (Wells and Pigliucci, 2000).

76

m c r« a « ln 9

V «le«H y —

Figure 2.14: Typical growth forms of freshwater macrophytes in slow to very fast flows
in shallow an deeper flowing waters (Dawson, 1988).

77

a)

b)

Figure 2.15: Two examples of heterophylly: (a) Cabomba caroliniana has dissected
submerged leaves and entire floating leaves, and (b) Potamogeton natans has linear
submerged leaves and ovate floating leaves. (From: http://plants.ifas.ufl.edu/cabaqu2.jpg;
Vic Ramey, University of Florida, and
http://www.miljolare.no/bilder/planter/ferskvann/makro/potamogeton_natans.jpg)

Chapter 3

78

The Effect of Macrophyte Morphology (Linear-bladed, Whorled, and Dissected) on
Fluid Dynamics
Abstract:
Fluid dynamic factors are important in defining the environments in which
macrophytes live, and have likely played a role in their evolution. For example,
freshwater macrophytes with different leaf morphologies (i.e., linear-bladed, whorled and
dissected) are common, but it is not known how these morphologies affect the flow
around them. An examination of the manner by which macrophytes affect their flow
environment was undertaken using flow visualization and image analysis of three
macrophyte species, Vallisneria americana (linear linear-bladed or ribbon-shaped),
Elodea canadensis (whorled), and Ceratophyllum demersum (dissected), and a circular
cylinder (physical model). The study was undertaken at five velocities ranging from ~1 1 1 cm s'\ using fluoroscein dye to visualize the flow patterns and digital recording to

capture the information. The flow patterns showed that at slower velocities the flow was
laminar as indicated by smooth dye streams, and stickiness as indicated by the loops of
dye attached to the leaf of Vallisneria, to the individual leaf whorls of Elodea, and to the
individual leaf segments of Ceratophyllum. At faster velocities, turbulence in the flow
was indicated by the generation of eddies, which were widely spaced downstream of the
cylinder, Vallisneria, and Elodea, and closely spaced downstream of Ceratophyllum.
The transition from laminar to turbulent flow occurred at lower velocities for the plants
than for a circular cylinder, with the exception of Ceratophyllum, thus indicating an
increase in local mixing. In addition, both the area of dye coverage and the concentration
of dye decreased at slower rates for the plants than for the cylinder, suggesting that plants
retain more fluid relative to the physical model. The consequence of these results

Chapter 3

79

indicated that there is both an increase in local flow around the plants and a retention of
fluid near the plants which provides more opportunity for nutrient exchange in the
boundary layers surrounding freshwater macrophytes. Therefore, it appears that aquatic
plants alter the flow in ways that maximize flow across leaves, which would enhance
nutrient exchange and thus macrophyte productivity.
Introduction:
Aquatic plants are important primary producers in aquatic ecosystems (Dodds and
Biggs, 2002), and are also ecosystem engineers in that they can (1) promote water quality
by providing physical protection for algal grazers; (2 ) compete with the algae for
nutrients; and (3) reduce water currents and allow suspended material to settle (Schutten
and Davy, 2000; Palmer et al., 2004; Wright et al., 2004). Interestingly, there are
freshwater macrophytes from different families that share common architecture and
habitat, and plants from the same families that differ in architecture suggesting that
genetics is probably not the only factor influencing plant morphology (Sculthorpe, 1967;
Schlichting, 1986; Ackerly et al., 2000; Givnish, 2002; Santamaria, 2002; Santamaria et
al., 2003; Dorken and Barrett, 2004). For example, Vallisneria and Elodea are in the
same family (Hydrocharitaceae monocotyledons) yet have very different architectures;
Vallisneria has long ribbon-shaped leaves, and Elodea has shorter whorled leaves.
Ceratophyllum (Ceratophylleaceae) and the submerged foliage of Cabomba
(Nymphaeaceae) in the dicotyledons have highly dissected whorled morphologies and are
in different families. Moreover, heterophylly (foliage plasticity) occurs in some species
of freshwater macrophytes whereby a plant may have different leaf shapes depending on
whether they are under, on, or above the water surface. Heterophylly occurs across

Chapter 3

80

distant taxa, which suggests that unrelated plants have adapted in a similar fashion to the
aquatic environment (Sculthorpe, 1967; Goliber and Feldman, 1990; Wells and Pigliucci,
2 0 0 0 ); adding to the argument that structural similarities occur due to adaptation to a

common fluid environment. The three common patterns of macrophyte morphology (i.e.,
linear-bladed, whorled, and dissected leaves) exist in different flow habitats ranging from
stagnant ponds to flowing streams (Dawson, 1988). Specifically, Vallisneria, Elodea and
Ceratophyllum all live in both still and flowing water (Cook et al., 1974). As such, it
would be appropriate to examine the nature of the factors that have potentially influenced
these morphologies, including fluid dynamic factors.
Fluid dynamics may play a large role in influencing macrophyte structure as fluid
dynamic factors are important in defining the microenvironments in which plants live
(Dawson, 1988). For example, plants living in energetic environments are exposed to
high drag forces that may cause damage to their leaves or blades, whereas in low energy
environments productivity may be limited by a thick boundary layer over which nutrients
are absorbed (Hurd et al., 1997; Schutten and Davy, 2000). Regardless of the
environmental conditions, a boundary layer begins at the surface of a boundary, where
the velocity is zero, and grows until the velocity of the boundary equals that of the
mainstream velocity (Vogel, 1994). Boundary layers are influenced largely by Reynolds
number (Re), which is the ratio of inertial to viscous forces (Vogel, 1994). The higher
the Re, the thinner the boundary layer, and the higher the shear rate (i.e., the faster the
fluid moves past a solid surface); and the lower the Re the thicker the boundary layer and
the lower the shear rate (Cheer and Koehl, 1987). If the Re is low (e.g., small, slowly
moving objects or low flow velocities), the flow is dominated by viscous forces, and is

Chapter 3

81

consequently smooth and orderly, encouraging the growth of a thick boundary layer
(Koehl, 1996). This thick boundary layer may lead to diffusional stresses (i.e., limited
mass transfer) due to the low flux of nutrients through the boundary layer to the leaf
surface compared to the rate at which the plants can utilize these nutrients.
Drag is also an important concept related to plants as it is the hydrodynamic force
that pulls a body in the direction opposite to fluid movement and in the case of boundary
layers, the force that slows the flow (Koehl, 1996). At low Re, drag is due to skin
friction, which is the viscous resistance of the fluid in the boundary layer around the body
being sheared as the fluid moves past the body. Consequently, the greater the surface
area, the greater the skin friction. At high Re, drag includes skin friction but is
dominated by pressure drag, which is the pressure difference across the body due to the
fluid dynamic separation or the formation of a wake on the downstream side of the body
(Vogel, 1987; Koehl, 1996). The drag force depends on the velocity of the fluid, the
frontal area and roughness of the object. Additionally, for a non-spherical object, the
shape and orientation of the object in relation to the flow are also important (Schutten and
Davy, 2000). For example, when plant stems bend downstream in response to faster
flow, their frontal area is reduced, lowering pressure drag. Long ribbon shaped leaves
can fold against the stem or each other, reducing skin friction (Ennos, 1999). It is thus
important to consider how fluid dynamic factors at low and high velocity effect the
morphological patterns observed in macrophytes.
This study will focus on the null hypothesis that plants with different leaf types
(linear-bladed, whorled and dissected) effect the downstream flow patterns in the same
manner. It is expected that Vallisneria (linear-bladed) will act in a manner similar to a

Chapter 3

82

physical model provided by a circular cylinder despite its flexibility, because of its
smootb-linear-bladed morphology. It is expected that Elodea (whorled) will also act
similar to the cylinder because of its rigid structure, and because the whorls of entire
leaves are likely spaced far enough apart to not have a significant effect on flow
separation. It is, however, expected that Ceratophyllum (dissected) will act more as a
mesh because of the high level of dissection in its leaves.
Materials and Methods:
The data were obtained from photographs and short videos of Fluorescein dye
moving past (I) the empty test section of the flume, (2) a circular cylinder (0.7 cm
diameter), (3) Vallisneria americana, (4) Elodea canadensis, and (5) Ceratophyllum
demersum (5 cm long sections). The empty flume was used to establish that the observed
patterns in the study were due to the subject-flow interaction. The circular cylinder was
used as a physical model for comparison as it is a well characterized fluid dynamic
system and a first approximation to an aquatic plant (Kundu, 1990; Coutanceau and
Defaye, 1991; Williamson, 1996; Palmer et al., 2004); E. canadensis was acquired from
Tabor Lake, Prince George, B.C., whereas V. americana and C. demersum were
purchased from Ward Scientific Supply (Rochester, NY, USA). Plants were maintained
in aquaria at -20 °C under natural light conditions, with potting soil (vermiculite
removed) as the substrate. Water velocities over the range of I-I I cms'* were used in
this study, which are similar to velocities used by Nepf and Koch (1999) and Leonard and
Luther (1995). These velocities are at the lower end of the range where the macrophytes
occur in the field (Wiegleb, 1984; French and Chambers, 1996; Elliot, 2000; Shi and

Chapter 3

83

Hughes, 2002; Wuest and Lorke, 2003), but are consistent with within-canopy flows
(Sand-Jensen and Mebus, 1996).
Flume
A recirculating flume of dimensions 19.2 cm (width) x 24.7 cm (height) x 170 cm
(length) was used for this study, with the test section being 1 2 0 cm downstream of the
flow straighteners (Figure 3.1). Velocity profiles were obtained using a 3-dimensional
Acoustic Doppler Velocimeter (ADV; SonTek; sampling at 25 Hz) for each mainstream
velocity examined during the experiments. These profiles show the logarithmic boundary
layer development, except at the faster velocities, where not enough data points were
collected in the logarithmic layer close to the bottom of the flume. The log portion of
these profiles was used to calculate the shear velocity (u* from m = — In— ; where u is

the velocity in the downstream direction, K is the von Karman constant (= 0.40), z is the
distance from the boundary, and zq is the roughness length (height): Table 3.1; Hoover
and Ackerman, 2001; Figure 3.2 a-e).
Photographs
Photographs or short videos (duration 40 seconds) were taken from both a
top and side perspective with a Nikon CoolPix995 (Melville, NY, USA) digital
camera placed approximately 60 cm directly across from the subject. The
camera was set with the ISO (sensitivity setting) at 400; the image adjustment at
lighten image; the shutter speed at 1/30; and the overall camera setting was at
shutter priority auto. All photographs were taken with a black cloth surrounding
the working section of the flume and the camera so that no external light could
enter the system.

Chapter 3

Dye (3 g/L sodium fluoroscein in distilled water) was released 3 cm
upstream of the subject and 3 cm vertically up from the base of the stem using a
peristaltic pump (mini-pump variable flow; slow flow, VWR, Canada; see Figure
3.1). Dye was released isokinetically, i.e., to match the mainstream velocity as
close as possible (Table 3.2). Dye was used to provide an indication of the
pattern of water flow around the subjects (i.e., physical model, and three plant
species).
Top perspective
A positioning device consisting of a modified syringe attached
perpendicularly to a Plexiglas plate was mounted to the side of the flume. The
subject (i.e., model or plant species) was placed in the top of the syringe and held
in place with mounting putty (Lepage, FUN-TAK) the putty was covered in
black tape in order to reduce reflectance (Figure 3.3). The subject was
positioned so that the dye would intercept in the centre, 3 cm up from the false
bottom of the flume. The lighting consisted of: (1) a 15 W florescent black light
(41 cm long, 2.5 cm diameter; General Electric; Burnsville, MN, USA) attached
approximately 15 cm above the flume, 15 cm downstream of the plant; and (2) a
6 V 15W white halogen mieroscope light (set at intensity # 6 ; VWR; Mississauga,

ON, Canada) that was placed on the stand direetly over top of the subject (Figure
3.1). A corrugated plastic sheet with a 1 cm wide and 17 cm long slit was placed
on the top of the flume to illuminate the flow behind the subject. It was assumed
that surface waves and buoyancy effect (i.e., gravity) were not likely significant
based on the depth of the positioning device below the surface (~ 6 - 8 cm below

84

Chapter 3

the surface depending on the velocity) and on the relatively short distance (~ 1 0
cm) that the dye travels horizontally. Based on these assumptions, the fluid
patterns in the empty test section were set as the base to which comparisons were
made (i.e., they were set as the control).
Side perspective
A similar set up was used for photographs in the side perspective. In this
case, the subject was placed in the hole drilled in a false bottom placed in the
flume and held in place with putty, such that the putty was flush with the false
bottom. The lighting scheme was the same as for the top view except that the
corrugated plastic sheet with the slit was not used (Figure 3.1).
Concentration Curve
A concentration curve was generated in order to convert from pixel value
in the photographs to the concentration of dye in the flow. This process was
undertaken separately for side and top perspectives because of the different
lighting schemes. The 100% dye concentration was set at 3g (sodium
fluorescein) /L (distilled water), and serial dilutions were made (i.e., 100-0.01%
[= 100, 10, 1, 0.1, 0.01%], 200-0.02%, 300-0.03% and 50-5%) omitting the
values above 100% in the analyses. The dye of a given concentration was placed
in a cuvette with black tape on the upstream and downstream sides and around
the cap to minimize leakage. The cuvette was placed upside down in the test
section of the flume and a cuvette containing 100% dye was placed 0.5 cm
upstream, and a cuvette containing 0% dye was placed 0.5 cm downstream of the
cuvette under investigation (Figure 3.4). Photographs of the cuvettes were made
using the same lighting scheme as was used to take the photos. The photos were

85

Chapter 3

imported into MATLAB version 6.5 (Mathworks, Natick, Massachusetts) for
conversion to gray scale and the background dye, determined as 69 in the side
perspective and 19 in the top perspective, was subtracted (see below). Since the
flume water was changed frequently (i.e., after every trial) the background dye
never exceeded the aforementioned pixel values. Five pixels were sampled from
each cuvette directly below the top of the cuvette, where the reflectance of the
plastic no longer influenced the dye. This technique was verified by sampling
pixels in the dye stream immediately downstream of the dye injector, where the
concentration of dye should be 100%. The results from both locations were
similar in side (186 ± 0 [mean ± standard error], n = 1 0 where a pixel value of
186 = 100% concentration) and top perspectives (89 ± 3, n = 10 where a pixel
value of 8 8 = 100% concentration). Lastly, the positions of the 100% and 0%
dye cuvettes were changed to determine whether the lighting varied across the
test section, but no differences (186 ± 0 , n = 1 0 for 1 0 0 % concentration, and 0 ±
0 , n = 1 0 for 0 % concentration) were detected.

The concentration curve (Figure 3.5) was non linear and was, therefore,
divided into two portions: (1) an exponential portion for 0.01-5% dye plotted
using a log-log transform; and (2 ) a linear portion from 1 0 - 1 0 0 % dye data
(Figure 3.6 a-d). The 10% dye concentration was used as the demarcation
between the two portions of the curve because that is where the curve changed in
the original plot (Figure 3.5).
Pixel values were grouped into fourteen bins of non-uniform size based
on the following equation, (((actual pixel value - estimated pixel v a lu e)/ actual pixel value)

86

Chapter 3

X 100%), where actual pixel is the value from the cuvette photographs, and
estimated pixel is the pixel value determined from the concentration curve. The
bin sizes are as follows: 0, 0.001-0.050, 0.051-0.150, 0.151-0.250, 0.251-0.500,
0.501-1.000, 1.001-2.000, 2.001-5.000, 5.001-10.000, 10.001- 20.000, 20.00140.000, 40.001-60.000, 60.001-80.000, and 80.001-100. The reason for non­
uniformity in bin size was to minimize the error in estimation. In this case the
degree of error accepted was set differently for the non-linear (< 1 0 % dye
concentration) and linear (> 1 0 % dye concentration) portions of the
concentration curve. The errors were 27% and 12 % for non-linear and linear
portions, respectively for the side perspective and 25% and 13% for the top
perspective. A mock photograph of known area was examined to test this
binning method (Figure 3.7). The distribution of colour within the photograph
was set as white = 23.6%; gray = 27.7 %; black = 48.7 %. The distribution
determined using the MATLAB based system was white = 24.3 %; gray = 27.0
%; black = 48.7%. These results indicate that the binning method appears to
work well and would be appropriate to apply to the dye patterns in the
photographs and videos. Since it is known that 100% dye concentration is 3g/L,
the actual dye concentration was calculated through cross multiplication and
used in further analysis.
Image Analvsis
A MATLAB program (Image Analysis Toolkit - version 6.5.0 Release
13.0.1) was used to edit the photos and to extract the digital information in the
flow around the subject of interest (e.g. cylinder or plant) (Box 3.1). Before the

87

Chapter 3

photos were processed in MATLAB, the subject was removed from the picture in
Adobe Photoshop (Version 5.0 limited edition; Adobe system, San Jose,
California) so that it would not be included in the analysis. The photos were
converted to gray scale so that the output was a two dimensional array
corresponding to the spatial location and pixel intensity (0 - 255). The
brightness and intensity of the image was then adjusted using MATLAB routines
to maximize visibility and any background level of dye (i.e., the dye in the flow
from an earlier trial) was subtracted from the image (see box 3.1). Four domains
of interest were identified: ( 1 ) upstream of the subject in the side perspective ( 6
cm high X 2 cm wide), (2) downstream of the subject in the side perspective ( 6
cm high X 7 cm wide), (3) upstream of the subject in the top perspective (4 cm
wide X 2 cm long) and, (4) downstream of the subject in the top perspective (4
cm wide x 7 cm long). These domains were compared in the statistical analyses,
which will be discussed in a later section.
The area of dye coverage was defined as any area in the domain (i.e.,
upstream vs. downstream in top and side perspectives) containing dye (i.e., a
non-zero pixel after background correction). The concentration of dye in the
domain was determined from these same data by converting the pixel intensities
to dye concentration using the relationship described above (see Concentration
Curve). This procedure used the frequency distribution of the pixel intensities,
which were multiplied by the concentration at the midpoint of each frequency
bin (recall that 100% dye concentration is 3g/L). This analysis was repeated for

Chapter 3

89

each of the three photographs taken at a given velocity. The average and
standard deviation of the three observations are reported.
Acoustic Doppler Velocimeter (ADV; SonTek) Measurements
Velocity measurements in three dimensions were taken on a horizontal plane (i.e.,
top perspective) 3 cm above the bottom of the flume using an ADV to examine the flow
around: (1) the empty test section; (2) a circular cylinder; (3) Vallisneria americana; (4)
Elodea canadensis; and (5) Ceratophyllum demersum. Each set of measurements was
taken at a frequency of 25 Hz for 120 s in a 2 cm x 2 cm grid that was 8 cm wide and 16
cm long. The exception occurred in the vicinity of the subject where measurements were
taken at a 3 cm distance due to interference of the acoustic signal with the subject (Figure
3.8). These measurements were made at the same five flume velocities as the photos in
order to have quantitative verification of any patterns that may occur. The ADV also
gives the root mean square of the velocity, which can be divided by the average velocity
to give turbulence intensity. This data was not included in this thesis because the trend
was similar to that seen in the velocity contour plots, where there was an increase in
turbulence intensity directly behind the plant. In all cases (cylinder and plants) the
turbulence intensity was approximately 60% upstream and beside the subject, and
approximately 100% behind the test subject at the lowest velocity (1.3 cms'^). At the
fastest velocity ( 1 1 . 0 cms'*), the turbulence intensity ranged from 2 0 % upstream and
around the cyinder to 40% behind the cylinder, in Vallisneria the range was from 2025%, in Elodea from 30-50%, and in Ceratophyllum from 25-30%.
Plant Angles and Oscillations
The downstream deflection and rate of oscillation of the plants in the flow were
examined using the 40 s video recordings. The angle of deflection (angle from the

Chapter 3

90

vertical) of the subject was measured for the cylinder and the three plant morphologies.
Since the plants did not remain stationary (i.e., they oscillated), the angle of deflection
was measured at the mean position taken at the location where the moving plant tips were
most frequently (Stephan and Gutknecht, 2002), under the assumption that the plant stem
is a straight line as in most cases there was a slight curve in the stem. The oscillations
were counted as the number of times the plant rose and fell during the 40 s recording.
With Vallisneria, only one blade was examined; the one that was parallel to the flow in
the downstream position. Additionally, the angle of the leaf from the stem was measured
for Ceratophyllum for the leaves in both the upstream and downstream position, from the
base of the plant to the apex.
Statistical Analysis
Effect of Velocitv on Dve Coverage and Concentration
To determine whether increasing velocity affected the area of dye coverage and/or
the concentration of dye in that area, the average and standard deviation in each domain
were determined for three images taken at each velocity. The resultant values (area or
concentration) were then regressed on the test velocity and stem Reynolds number based
on the diameter of the subject (i.e..

Re = —

, where u is the velocity, 1 is the diameter

V

of the plant stem and whorl, and o is the kinematic velocity; Vogel, 1994). The null
hypothesis examined was that the slope of the regression would equal zero. In other
words there is no relationship in area of dye coverage or concentration of dye with
increasing velocity.
Effect of Location and Perspective on the Dve Coverage and Concentration

Chapter 3

91

An analysis of covariance (ANCOVA) was used to test the null hypothesis that
there is no difference in the upstream and downstream flow patterns in both the side and
top perspectives for the different subjects. Area was normalized by using the relative
frequency (# pixels of dye/ total # pixels x 1 0 0 %) of dye in the ealeulation of the area of
dye coverage. In this case, the area of dye coverage was used as the dependent variable,
the domains (i.e., side or top perspeetive, upstream or downstream) were the categorical
factors, and velocity was the continuous predictor (the covariate). Comparisons between
side and top views were make within the same test section. A Fisher least significant
difference (LSD) test was used for Post Hoc pair wise eomparisons. The same analysis
was used to examine the eoneentration of dye in the specified area.
Effect of Morphologv on the Rate of Oseillation
ANCOVAs were condueted to test the null hypothesis that the rate of stem
oscillation did not differ among plant morphologies. The oseillation rate was used as the
dependent variable, the different subjects were the categorical factors, and velocity was
the continuous predictor. A Fisher LSD test was used for Post Hoe eomparisons.
Effect of Velocitv and Node Position on the Leaf Angles of Ceratophyllum
ANCOVAs were used to test the null hypothesis that the position of the leaf on
the stem (i.e., from base to apex) had no effect on the angle of the leaf. The angle of the
Ceratophyllum leaf from the stem was used as the dependent variable, the node position
on the stem was the categorical factor, and velocity was the continuous predictor. A
Fisher LSD test was used for Post Hoe eomparisons.
Effect of Location on Velocitv

Chapter 3

92

ANCOVAs were used to test the null hypothesis that the pattern of velocity
upstream of the subject is the same as the pattern of velocity downstream of the subject.
This was done by grouping together the upstream ADV velocity measurements (14
points) and the downstream ADV velocity measurements (30 points). The average
measured velocity (of each point) was used as the dependent variable, the position
(upstream or downstream) of the measurements was the categorical factor, and test
velocity was the continuous predictor (the covariate). A Fisher LSD test was used for
Post Hoc pairwise comparisons. Since the area where the ADV velocity measurements
were taken was larger then the area used in the image analysis, only the measures of
velocity that were in the same area used in photos from the top perspective were used
(i.e., 4 cm wide and 2 cm long upstream, and 4 cm wide by 7 cm long downstream.
Results:
Empty flume
The flow, as indicated by the dye streaklines was similar in both side and top
perspectives (i.e., left and right panels, respectively (Figure 3.9a-j). The dye streaks were
approximately linear at the two lowest velocities (1.3 and 2.0 cm s'\ Figure 3.9a-d), but
became “wavy” at the higher velocities (5.0, 8.4, and 11.0 cms"\ Figure 3.9e-j). In
addition the dye streaks were wider (had less contrast) at the higher velocities indicating
relatively fast flows (e.g. Figure 3.9i-j). In terms of the area of dye coverage, there was
little evidence of an effect of velocity in the “upstream” or “downstream” (defined
arbitrarily here as there was no subject placed in the flow) domains (Figure 3.10a-d). The
concentration of dye imaged in the domains tended to decrease with increasing velocity,
and this was significant in three of the four domains and was on the order of -O.IgL
Vcms'^ (Figure 3.10f-h).

Chapter 3

93

In terms of the area of dye coverage in a specified domain, the ANCOVA showed
that there was a significant difference between the flow upstream and downstream of the
empty test section (p < 0 .0 0 1 ), but not between the side and top perspectives (p = 0.082;
Table 3.3a), suggesting that surface waves and buoyancy do not have a significant effect
on the side vs. top fluid patterns. A significant interaction (p = 0.047) was found for the
side versus top perspectives by the upstream and downstream position. The post hoc
comparisons for the empty test section support this, showing that significant differences
occurred between all domains (p < 0.003), with the exception of the side and top
perspectives upstream of the test section (p = 0.81: Table 3.4a).
In terms of the concentration of dye in a specified domain, the ANCOVA showed
that there was a significant difference between the upstream and downstream positions (p
= 0.017), but not between side and top views (p = 0.95; Table 3.3a), supporting the
assumption that surface waves and buoyancy do not have a significant effect on the side
vs. top fluid patterns. Post hoc comparisons agreed, whereby significant differences
occurred between the upstream and downstream position in the side perspective (p =
0 .0 1 1 ), between the side perspective upstream and the top perspective downstream (p =

0.034) and between the top perspective upstream and the side perspective downstream (p
= 0.042; Table 3.5a).
Cylinder
The dye streakline upstream of the circular cylinder did not appear to change with
increasing velocity, in that it remained in a straight stream in both side and top
perspective (Figure 3.11). However, downstream of the cylinder, there was laminar flow,
with loops of dye that indicate “stickiness”, and a large area of recirculation at the slower
velocities. The transition between laminar and turbulent flow as indicated by the onset of

Chapter 3

94

eddies occurred between 5.0 cms'^ and 8.4 cms'^ (Re ~ 350-590). At the faster velocities,
the flow was turbulent, and as such less sticky, and the recirculation zone was smaller.
The dye streaklines downstream in the top perspective revealed a classic attached eddy at
1.3 cms'* (Re ~ 90; Figure 3.11b), a von Karman vortex street at 2 cms'^ (Re ~ 140;
Figure 3.lid ), and more turbulent structures at higher velocities (Figure 3.11f,h,j).
The image analysis indicated that there was no relationship between the area of
dye coverage and velocity in the upstream flow (p = 0.19 and 0.28; Figure 3.12a,b).
However, there was a significant increase, on the order of 0.5 cm^/cms'\ in the dye
coverage downstream in the side perspective with velocity (p = 0.018; Figure 3.12c), and
a tendency for increase in the top perspective (p = 0.41; Figure 3.12d). Additionally,
there was a tendency for the dye concentration to decrease with velocity in the upstream
and downstream domains (e.g., ~ -0.05 gL'Vcms'*), and this was significant in the case of
the downstream side perspective (p = 0.014; Figure 3.12g).
In terms of the area of dye coverage in a specified domain, the ANCOVA
indicated that there were significant differences upstream and downstream of the cylinder
(p < 0.001), but no differences occurred between the side and top perspectives (p = 0.52;
Table 3.3b). The post hoc comparisons concurred in that there were significant
differences (p < 0 .0 0 1 ) between the upstream and dowstream domains, but no significant
difference between side and top perspectives upstream of the cylinder (p = 0.94; Table
3.4b). Furthermore, there was no significant difference between the side and top
perspective downstream of the clinder (p = 0.28).
The ANCOVA of the concentration of dye indicated that there was no significant
difference between side and top perspectives (p = 0.39), nor between the upstream and

Chapter 3

95

downstream position of the cylinder (p = 0.22: Table 3.3b). Post hoc comparisons also
showed that domain had no significant effect on the concentration of dye (Table 3.5b).
Vallisneria americana
The dye streakline upstream of Vallisneria remained in a relatively straight stream
in both side and top perspeetive (Figure 3.13). The dye streaklines downstream showed a
large amount of recirculation under the leaf, and a high amount of stickiness on the leaf.
The flow was laminar at 1.3 cms"' (Re ~ 25) as indicated by the smooth dye streakline
(Figure 3.13a,b); the transition from laminar to turbulent occurred at about 2.0 cms"^ (Re
~ 40) when eddies began to detach (Figure 3.13c,d); von Karman vortices were shed at
5.0 cms ' (Re ~ 100; Figure 3.13f); and more widely spaced eddies with less stickiness
and recirculation were observed at the higher velocities (Figure 3.13g-j).
The image analysis indicated that there was no relationship between the area of
dye coverage and velocity in both the upstream (p = 0.072) and downstream (p = 0.69)
flow in the top perspective (Figure 3.14b,d). However, there were significant increases
in the area of dye coverage with velocity in both the upstream (p = 0.034) and
downstream (p = 0.014) flow in the side perspective, with slopes of 0.03 cm^/cms'^ and
0.2 cm^/cms"\ respectively (Figure 3.14a,c). In addition, there was a significant decrease
of -0.05 gL'Vcms"' in the dye concentration with velocity in the side perspective
upstream (p = 0.042), although a non-significant increase in the top perspective upstream
(p = 0.64; Figure 3.14e,f). There was also a tendency for the dye concentration to
decrease with velocity in the up and downstream domains (Figure 3.14g,h).
The ANCOVA of the area of dye coverage in a specified domain, revealed that
there was a significant difference in the flow upstream and downstream of Vallisneria (p
< 0.001), but not between the side and top perspectives (p = 0.37; Table 3.3c). The post

Chapter 3

96

hoc comparisons indicated the same whereby, there were significant differences between

the upstream and dowstream domains in the top and side perspectives (p < 0 .0 0 1 ), and no
significant difference in the side and top perspectives upstream (p = 0.95) or downstream
of Vallisneria (p = 0.26; Table 3.4c).
The ANCOVA of the concentration of dye showed that there was no significant
difference between side and top view (p = 0.84), nor between the up and downstream
position of Vallisneria (p = 0.64: Table 3.3b). Post hoc comparisons showed that no
significant effect occurred between any of the domains (Table 3.5c).
Elodea canadensis
The dye streakline upstream of Elodea was reasonably straight up to a velocity of
8.4 cms'^ (Re ~ 840) when some waviness was noted (Figure 3.15). The dye streaklines
downstream were attached to the individual whorls of leaves (i.e., “sticky”), were in
filamentous loops at the lower velocities and recirculation was present (Figure 3.14a-d).
The transition from laminar to turbulent occurred at ~ 5.0 cms ' (Re ~ 500) when the dye
streak was no longer smooth and eddies began to form (Figure 3.15e,f ); von Karman
vortices were shed at 8.4 cms'^ (Re ~ 840; Figure 3.15h); and turbulent structures
occurred at 11.0 cms'^ (Re ~ 1100) as indicated by the widely spaced eddies, less
stickiness on leaves and lower recirculation (Figure 3.15i-j).
The image analysis indicated that there was no relationship between the area of
dye coverage and velocity in the upstream flow in both the side and top perspective
(Figure 3.16a,b), nor in the side perspective downstream (Figure 3.16c). However, there
was a significant increase in the area of dye coverage with velocity downstream in the top
perspective (p = 0.026) of ~ 0.10 cm^/cms'* (Figure 3.16d). There was also a tendency
for the dye concentration to decrease with velocity in the up and downstream domains.

Chapter 3

97

and this was significant in the case of the downstream side (p = 0.049) and top
perspectives (p = 0.046; Figure 3.16g,h).

The ANCOVA of the area of dye coverage showed that there was a significant
difference between the flow upstream and downstream of Elodea (p < 0.001), but no
significant difference between the side and top perspectives (p = 0.26; Table 3.3d). The
post hoc comparisons showed that there were significant differences between the
upstream and dowstream domains in the side and top perspective (p < 0.001). No
significant differences were detected in the side and top perspectives upstream (p = 0.58),
or downstream of Elodea (p = 0.26; Table 3.4d), but a significant interaction (p = 0.043)
was found between the side versus top perspective by upstream versus downstream
comparisons (Table 3.3d).
The results of the ANCOVA of the concentration of dye showed that there was no
significant difference between side and top view (p = 0.29), or between the up and
downstream position (p = 0.30: Table 3.3d). Post hoc comparisons showed that domain
had no significant effect on the concentration of dye (Table 3.5d).
Ceratophyllum demersum
The dye streakline upstream of Ceratophyllum was relatively straight although
some waviness was noted at 5 cms'^ (Figure 3.17). The dye streaklines downstream of
Ceratophyllum showed a high amount of “stickiness” to the individual leaf segments due
to laminar flow at 1.3 cms'^ (Re ~ 300), along with a large amount of recirculation
(Figure 3.17a,b). The flow became transitional between 5.0 and 8.4 cms'* (Re ~ 12002000) where the loops of dye could no longer be detected (Figure 3.17c-f). Turbulent
structures were evident at the higher velocities as indicated by closely spaced eddies, and
a reduced amount of stickiness and recirculation (Figure 3.17g-j).

Chapter 3

98

There was no relationship between the area of dye coverage and velocity in the
upstream flow in the side and top perspectives (Figure 3.18a,b). However, there was a
significant decrease in the area of dye coverage with velocity downstream in the top
perspective (p = 0.028) on the order of -0 . 1 2 cm^/cms'\ and a similar trend in side
perspective (p = 0.31; Figure 3.18d). The dye concentration tended to decrease with
velocity in all perspectives, but these were not significant (Figure 3.18e-h).
Results of the ANCOVA of the area of dye coverage indicated that there was a
significant difference between the flow upstream and downstream of Ceratophyllum (p <
0.001), and only marginal significance between the side and top perspectives (p = 0.057;
Table 3.3e). A significant interaction (p = 0.047) was found in the side versus top
perspective by upstream versus downstream position. The pairwise comparisons
revealed signficant differences between all combinations of perspectives with the
exception of top versus side upstream (Table 3.4e).
The ANCOVA of the concentration of dye showed that there was no significant
difference between side and top perspective (p = 0.25), nor in the up and downstream
position (p = 0.83: Table 3.3e). Post hoc comparisons showed that domain had no
significant effect on the concentration of dye (Table 3.5e).
Comparison of the Three Plant Morphologies and the Theoretical Model
A direct comparison of the results should reveal whether there are differences
among the three plant morphologies and the physical model. Firstly, the velocity at
which the transition from laminar (smooth streakline) to turbulent (eddies) flow occurs
was lower for the plants, with the exception of Ceratophyllum, than for the cylinder
(Figure 3.19a). The transition occurred in Vallisneria (linear-bladed) at ~2 cms"\ in

Chapter 3

99

Elodea (whorled) between ~2-5 cms‘\ and in Ceratophyllum (dissected) between -5-8.4
cms'V It should be noted that the use of Re does not change the pattern of observation
reported here. Specifically the Re for the cylinder ranged from 90-770, Vallisneria from
25-220, Elodea from 130-110, and Ceratophyllum from 300-2600. Secondly, in all
cases, the rate of change in the area of dye coverage with velocity was lower for the
plants than for the cylinder (Figure 3.19b). More specifically, the rate of change
increased significantly for the cylinder at a rate of -0.5 cm^/cms’', as it did for
Vallisneria at -0.2 cm^/cms'\ There was no significant trend for Elodea (- -0.0004
cmVcms’'), and the rate for Ceratophyllum decreased non-significantly at - -0.2
cm^/cms'\ Thirdly, the rate of change in the concentration of dye with velocity also
decreased at a slower rate for the plants then for the physical model (Figure 3.19c). The
concentration of dye downstream of the circular cylinder decreased significantly at a rate
of - -0.05 gL'Vcms'*. For Vallisneria, the trend was non-significant with the decrease
being on the order of - -0.07 gL'Vcms'\ The rate for Elodea decreased significantly on
the order of - -0.06 gUVcms '. Although the rate was similar for Ceratophyllum (- 0.06 gL'Vcms"^), it was non-significant. The non-significant trends are likely due to a
higher degree of variance.
ANCOVA s were used to determine if there were differences among the cylinder
and the linear-bladed, whorled, and dissected leaf morphologies, when adjusted for a
common mean velocity and a common regression line. Plant morphology (p < 0.001:
Table 3.6a), side versus top perspective (p = 0.044), and upstream versus downstream
position (p < 0 .0 0 1 ) were found to have a significant effect on the area of the domain
covered in dye, whereas there was no significant effect of velocity (p = 0.083). More

Chapter 3

100

Specifically, post hoc comparisons revealed that there were no significant differences
between subjects in the upstream direction from both the side and top perspectives (Table
3.7a,b). This indicates that the upstream flow was similar in all cases. There were
significant differences in the side perspective, downstream of the subjects between: ( 1 )
the cylinder and the empty test section (p = 0.007: Table 3.7c); (2) Elodea and
Vallisneria (p = 0.043); (3) the empty test section and Elodea (p = 0.002); and (4) a
marginally significant difference (p = 0.059) between Ceratophyllum and the empty test
section. There were also significant differences in the top perspective, downstream of the
subjects between: (1) all subjects and the empty test section (p < 0.009: Table 3.7d); (2)
the cylinder and Ceratophyllum (p = 0.036); and (3) Vallisneria and Ceratophyllum (p =
0.010). There were also significant interactions between the plant and the upstream and
downstream position (p = 0.003), and between the side and top perspectives and the
upstream and downstream positions (p = 0.014; Table 3.6).
The concentration of dye was affected significantly by velocity (p < 0.001: Table
3.6b). However, neither plant morphology (p = 0.13), side versus top perspective (p =
0.17), nor upstream versus downstream position (p = 0.17) had a significant effect on the
concentration of dye. Post hoc comparisons revealed that there was no significant
difference upstream of the subjects in the side and top perspectives (Table 3.8a,b). The
only significant differences that occurred in the downstream position was in the side
perspective between the empty test section and the cylinder (p = 0.005), Elodea (p =
0.025), and Ceratophyllum (p = 0.038; Table 3.8c). Marginally significant differences
were observed in the top downstream perspective for the circular cylinder and
Ceratophyllum (p = 0.055) and the empty test section (p = 0.052).

Chapter 3

101

Velocity Measurements
ADV measurements were taken in a horizontal slice and velocities were recorded
in u (streamwise), v (cross-stream), and w (vertical) components, though statistics were
only performed for the u and w velocity components. The contour plots of the u velocity
component showed a pattern of decreasing velocity behind the subject (Figures 3.203.24a,c). The patterns in both the streamwise (u) and vertical (w) velocity components in
the empty test section showed a relatively uniform pattern, although there was more
variance with increasing velocity (Figure 3.20a-d). The velocity pattern downstream of
the circular cylinder showed that there was an area of reduced velocity which, diminished
in size with increasing velocity (Figure 3.21a,c). In the vertical component, there was
downward velocity (i.e., negative velocity) directly downstream of the cylinder at the
slower velocity (Figure 3.21b), and this region moved ~ 5 cm downstream of the cylinder
at the faster velocity (Figure 3.2Id). The velocity patterns in the streamwise component
(u) of Vallisneria showed a small area of reduced flow downstream of the plant at the
slower velocity (Figure 3.22a), but this pattern disappeared as velocity increased (Figure
3.22c). Similarly, in the vertical component (w), the flow moved upwards directly
downstream of the plant at the slower velocity (Figure 3.22b), with no obvious pattern
occurring at the fastest velocity (Figure 3.22d). Elodea showed a pattern of reduced flow
in the streamwise component (u), though this area of reduced flow was much larger ( - 1 0
cm vs. 5 cm) at the faster velocities (Figure 3.23a,c). In terms of w, the flow moved
upwards downstream of the plant at the slow velocity (Figure 3.23b), but no obvious
trend was noted as velocity increased (Figure 3.23d). Finally, Ceratophyllum showed an
area of reduced flow directly downstream the plant at the slower velocity in the

Chapter 3

102

streamwise component (u) (Figure 3.24a), but the flow remained fairly uniform as
velocity increased (Figure 3.24c). In the vertical velocity component (w), the flow was
fairly uniform at the slower velocity (Figure 3.24b), but began to move upwards directly
downstream the plant as velocity increased (Figure 3.24d).
An ANCOVA of the velocity data showed that there was a significant effect of
test subject on the streamwise velocity (u) at the different test velocities (p < 0.001; Table
3.9a), as well as on the upstream and downstream position (p = 0.023), however, there
was no effect between plant morphologies (p = 0.47). In the vertical velocity (w), there
was also a significant effect on the velocity contours at the different test velocities (p =
0.014: Table 3.9b). Additionally, there was a significant difference in velocity between
the different plant morphologies (p < 0 .0 0 1 ), and between the upstream and downstream
positions (p < 0.001). Post hoc comparisons of the streamwise component (u) showed
that there were significant differences between downstream of the cylinder and of the
empty test section (p = 0.044: Table 3.10a), between the upstream and downstream
position of Elodea (p = 0.034), between upstream of Elodea and downstream of
Vallisneria (p = 0.010), between downstream of Elodea and the empty test section (p =
0.014), between up and downstream position of Vallisneria (p = 0.034), between
downstream of Vallisneria and upstream of the empty test section (p = 0.046), and
between downstream of Vallisneria and the empty test section (p = 0.0021). An
evaluation of the vertical velocity component (w), indicated that there were significant
differences between upstream and downstream position of the cylinder (p < 0.001: Table
3.10b). Downstream of the cylinder was significantly different than all other plant and
position combinations (p < 0.002), as was upstream of Elodea (p < 0.008).

Chapter 3

103

Plant Response to Flow
The angle of deflection remained stationary at 0° for the circular cylinder (Figure
3.25a), however it increased with velocity for the plants. The deflection from the vertical
increased significantly for Elodea (p < 0.001: Figure 3.25c) and Ceratophyllum (p =
0.003; Figure 3.23d). The results for Vallisneria (p = 0.30; Figure 3.25b) were not
significant. An ANCOVA of these results indicated that velocity had a significant effect
on the angle of deflection (p < 0.001), as did the plant morphology (p < 0.001; Table
3.11a). Significant pairwise differences (p < 0.008: Table 3.11b) were found between all
combinations of subjects. To summarize, the rate of change in the angle of deflection
remained at zero for the cylinder due to its rigid architecture. For Vallisneria there was a
slight though non-significant increase, a larger and significant increase for Elodea, and a
larger and also significant increase for Ceratophyllum, thus demonstrating that the more
complex the plant architecture, the greater the rate of change in the angle of deflection
with increasing velocity (Figure 3.26a).
Oscillations were not observed for the circular cylinder (Figure 3.25e).
Significant or near significant increases in oscillation were observed for Ceratophyllum
(p = 0.018; Figure 3.25h), Vallisneria (p = 0.087; Figure 3.25f), and Elodea (p = 0.083;
Figure 3.25g) and the rate was on the order of 0.05 Hz/cms V The ANCOVA revealed
significant effects of velocity (p < 0.001: Table 3.12a) and plant morphology (p < 0.001).
Pairwise comparisons however, showed that all differences occurred between the
cylinder and the plants (p < 0.001: Table 3.12b), with no significant differences in
oscillation rate occurring between the different plant morphologies. As an overview, the
cylinder did not oscillate because of its rigid structure, the three plant morphologies

Chapter 3

104

however, showed a similar trend of an increasing rate of oscillation with increasing
velocity, which was significant only for Ceratophyllum (Figure 3.26b).
Discussion:
Macrophvte-Flow Interaction
The transition from laminar to turbulent flow (smooth dye streaks to eddies)
occurred at lower velocities for the plants, with the exception of Ceratophyllum, than for
the cylinder. This is interesting because Hurd and Stevens (1997) documented that the
blades of marine algae induce the transition to turbulence at lower velocities than a flat
plate, which was their physical model. Additionally, the transition for Ceratophyllum
occurred at approximately the same velocity (> 5 cms'^) as for the coarsely branched
algae Gelidium coulteri (Hurd and Stevens, 1997). The transition to turbulent flow is
important because it lead to a reduction in the thickness of the diffusional sublayer on the
leaf surface, thus enhancing local mixing and nutrient delivery (Anderson and Charters,
1982; Borchardt, 1994; Hurd et al., 1996; Nepf, 1999). Additionally, both the area of dye
coverage and the concentration of dye decreased at slower rates for the plants than for the
circular cylinder, which demonstrates that plants retain more fluid, thus increasing the
opportunity for exchange between the water column and the leaf surface. This would
indicate that both fluid retention and mixing occur around the plants. This would provide
the opportunity for nutrient uptake by the plants and replenishment of nutrients from the
surrounding fluid. In other words, the recirculation on the downstream side of the plant
increases the opportunity for the well mixed fluid to come in contact with the plant
surface. Therefore it appears that aquatic plants alter the flow in ways that maximize
flow across leaves, which would enhance nutrient exchange.

Chapter 3

105

In this study, a circular cylinder was used as a physical model because it is a
simple form, and the downstream flow patterns have been well documented (e.g., Kundu,
1990). The similarities among the plants (with the exception of Ceratophyllum) likely
occurred because the Reynolds numbers examined here (cylinder (Re = 90-770),
Vallisneria (Re = 25-220), Elodea (Re = 130-1100), and Ceratophyllum (300-2600))
were above the point (with the exception of Vallisneria) where attached vortices are shed
(Re > 80; Kundu, 1990), and below the point where the wake is completely turbulent (Re
< 5000; Kundu, 1990). The wake remained laminar for all Re examined here, however,
changes occurred in the downstream eddies which began as small recirculating vortices,
and grew with increasing Re and eventually became unstable and detached (von Karman
vortex street; Figure 3.27). As such, the experimental results coincided with the
theoretical expectations in most cases. Specifically, it was expected that Vallisneria
would act like a cylinder because of its uniform structure (i.e., no branches). This was
generally the case except that it retained more fluid than the cylinder, perhaps due to the
inclined orientation. It was expected that Elodea would act similar to the cylinder
because of its rigid cylinder-like structure. This was however, not the case as the leaf
whorls trapped the dye and the regression analysis indicated that Elodea was more
similar to Ceratophyllum. Ceratophyllum was expected to act more as a mesh, because
of its high level of dissection which was predicted to break up the vertical circulation
behind the plant, similar to a mesh (Nowell and Jumars, 1984). This is an important
distinction because turbulence is dissipated by the resistance of the mesh, and as such is
straightened, as has been demonstrated with the red alga Gelidium nudifrons (Anderson
and Charters, 1982), and with Gelidium coulteri (Hurd and Stevens, 1997). Additionally,

Chapter 3

106

as Re increases, vortices form closer to the branch (Kundu, 1990), leading to less
entrainment of the dye because the recirculation zone is smaller. Hence, if features on
the plant generate small scale turbulence it could enhance mass and momentum transfer,
and thus productivity. However, it has also been suggested that that since mesh-like
structures slow fluid down and dampen turbulence, there could be the potential for
diffusion-limited productivity to be greater in branched rather than linear-bladed
morphologies (Hurd and Stevens, 1997). This is, however unlikely considering that it has
been shown that the critical velocity where limitations change from diffusional to
mechanical stress is between 2 and 6 cms'' (Jumars et al., 2001; Hurd et al., 1996; Koch,
1993; Koch, 1994). This is consistent with this study as the transition from laminar to
turbulent flow occurred at approximately 5 cms"\
It was reasonable to expect differences in area of dye coverage among plant
morphologies (Vallisneria > Elodea > Ceratophyllum) because leaves can act as edges to
which dye can attach, and flow separation occurs (Coutanceau and Defaye, 1991);
separation would increase the dispersion of the dye. Separation points occur when the
flow stops moving up the surface of the plant, and starts moving in the downstream
direction (Vogel, 1994). Stevens and Hurd (1997) noted that under almost any field
conditions in the marine environment, the pressure gradients at leading edges of
macrophyte leaves will be sufficient to cause separation in the flow. This process of
separation, whereby the separated shear layer attaches on the solid boundary, forms a
closed recirculating zone (Kiya et al., 2000), has been shown to enhance mass-transfer
(Schwinge et al., 2002). This area of recirculation would be much larger for Vallisneria
than for Elodea and Ceratophyllum because the size and shape of recirculation is

Chapter 3

107

dependent on how the leaves are arranged, their diameter and length (Schwinge et al.,
2002), and the boundary layer thickness (Lozee and Wetzel, 1993). This is demonstrated
by flow through a mesh, which reduces highly turbulent flow structures to smaller eddies
(Nowell and Jumars, 1984; Sand-Jensen and Pedersen, 1999) because of the thinness of
the structure and the small size of the openings.
The rate of change in the concentration of dye with change in velocity decreased
in all cases, but less so for the plants than for the cylinder, which indicates that more fluid
is being retained by the plants. It is interesting that the area of dye coverage is affected
by the different plant morphologies, but that the concentration of dye downstream of the
plant is not. In other words, the dye dispersed differently (i.e., area changes) according to
the plant shape, but the same amount of dye is retained. Comparatively, Hurd et al.
(1996) showed that blade morphology (i.e., undulate versus smooth) affected the flow
patterns, with the undulations generating recirculating eddies, and the smooth blades
acting like plates. However, despite differences in flow patterns, nitrogen uptake was
similar between the two different blade morphologies. This supports the previous
statement demonstrating that that freshwater macrophytes of different morphologies have
different ways of interacting with the flow, while retaining the same amount of fluid.
In addition to plant morphology, the angle of the plant in the flow is important as
an increase in the angle of deflection results in a decrease in the drag forces acting on the
macrophyte (Dawson and Robinson, 1984). Consequently, plants living in faster moving
water have leaves with low branching angles, which produce a more streamlined surface.
In contrast plants living in slower moving water have leaves that stick out more into the
flow presenting a rough surface which, increases the drag coefficient, but results in the

Chapter 3

108

production of small vortices behind the leaves that increase the supply of carbon dioxide
and nutrients, thereby increasing the rate of photosynthesis (Ennos, 1999; Sand-Jensen
and Pedersen, 1999; Schutten and Davy, 2000). Moreover, it is well documented that
smooth, flexible stems are best able to reduce drag (Armstrong, 1989; Johnson, 2001),
especially when the angle of deflection is greater than 30°, the plant is more streamlined
and the pressure drop around the stem is reduced. When deflected almost horizontally
(as seen with C eratophyllum and V allisneria) the leaf can block vertical exchange flow
(Nepf and Koch, 1999). This could deprive the leaves higher up in a canopy of nutrientenriched water if the flow was prevented from travelling vertically. However, if the flow
is not blocked, secondary flow may promote direct transport between the sediments and
the water column and thus lessen diffusive boundary layer limitations on nutrient and
inorganic carbon exchange (Nepf and Koch, 1999).
Furthermore, the rate of change in the angle of deflection with change in velocity
increased with increasing complexity in architecture (i.e., V allisneria < E lodea <
Ceratophyllum ). This is likely because V allisneria's architecture is already deflected

from the vertical, E lodea is relatively rigid with whorls of leaves, and Ceratophyllum is
relatively flexible with highly dissected leaves exposed to the flow. It is also interesting
to note that C eratophyllum , which was the most compliant plant retained the least dye
and Vallisneria, the plant whose compliance changed the least, retained the most dye
(Figure 3.19). This relates back to the argument that the size of the recirculation zone is
dependent on the size, shape, and orientation of the leaf (Schwinge et al., 2002). There
were however, no significant differences among the different plant morphologies and the
rate of change in the frequency of oscillations with velocity. Oscillations are due to

Chapter 3

109

phase differences between lift and the shedding frequency of vortices, and are not found
for circular cylinders (Norberg, 2003). These plant movements are important because
they cause a dampening of the flow which in turn causes a dissipation of mechanical
energy, which increases the effectiveness of energy transfer (Bruchert et al., 2003).
Additionally, oscillations can noticeably change the formation of vortex shedding
patterns even with cylinders (Williamson, 1996), potentially increasing local mixing.
Local mixing can also be increased because the waving of plants (i.e., monami) has been
shown to cause in increase in the vertical transport of momentum into the canopy
(Ackerman and Okubo, 2003; Ghisalberti and Nepf, 2002).
Finally, the technique of isokinetic dye injection appears to be valid in that
differences in upstream flows were not detected. Changes present in the relative
downstream position of the empty test section were due to the expansion of the dye
plume, which increased with distance from the source (i.e., dye injector; Mercier and
Jaluria, 1999; Macdonald et al., 2002; Hara and Kato, 2004).
Implications of the Macrophyte-FIow Interaction
The significance of these findings relate to issues of macrophyte productivity
(e.g., photosynthesis and nutrient uptake), sediment capture, and the evolutionary
question about form and function relationships, that has previously been examined in
seaweeds (Norton et al., 1982). Light is the limiting factor in determining the maximum
depth at which plants occur in a given body of water (Riemer, 1993), and since it is
known that both low (Hurd and Stevens, 1997; Schutten and Davy, 2000) and high
(Koch, 1993) flows affect photon capture, macrophytes must adapt to maximize
photosynthesis, and thus primary productivity. At low flow, a thick diffusive boundary

Chapter 3

110

layer occurs which, leads to a decrease in the amount of carbon availability, and thus a
reduction in photosynthesis (Koch, 1993;1994). At high flow, photosynthesis can be
limited by the washing away of extracellular enzymes that catalyze the reaction between
bicarbonate and carbon dioxide (Koch, 1994), and by breakage of the photosynthetically
active parts of the plant (Koch, 1993; Stewart and Carpenter, 2003). The question of
form and function relationships relates to the structural similarities of the plants on fluid
dynamics. For example, linear-bladed morphologies are better suited to higher flows
because they are better able to comply to the flow (Gutierrez and Fernandez, 1992) and
thus are damaged at higher velocities than branched morphologies (Sheath and
Hambrook, 1988). Conversely, morphologies with small scale roughness are able to
induce turbulence at lower velocities thus enhancing local mixing and preventing the
depletion of nutrients. For example, undulate margins have been shown to take up
labeled carbon at a faster rate than smooth blades at the same velocity (Armstrong, 1989).
This is congruent with the current study whereby Vallisneria interacted the least with the
flow because of its high angle of deflection, Ceratophyllum broke the flow up into
smaller turbulence structures, and Elodea affected the flow in a fashion between these
two extremes. Thus, the interaction between macrophyte morphology and fluid dynamics
has implications for the productivity of the individual plant and the ecosystem.
Quantitative similarities in the concentration of dye among plant morphologies
may result from different flow patterns that maximize contact of flow with the leaves,
and as such provide light and nutrients to the leaf surface. But, as mentioned above, a
morphology that increases nutrient uptake may not be the same morphology that
increases photosynthesis, as high levels of dissection may cause self-shading (Norton et

Chapter 3

111

a l, 1982). As well, it has been documented that linear-bladed seaweeds have higher rates
of photosynthesis than branched morphologies (Irwin and Davenport, 2002). This
suggests that Vallisneria should have better photosynthetic abilities than the whorled and
dissected morphologies, and Ceratophyllum should have better nutrient uptake than the
linear-bladed and whorled morphologies.
Leaf development in aquatic macrophytes is a dynamic process which,
emphasizes the close relationship between environment and leaf structure (Rascio, 2002).
One classic example of this is found in a number of seaweeds where high drag
morphologies (e.g., rugosity and undulation) are found in low energy environments and
low drag morphologies (e.g., smooth, straight blades) are found in high energy
environments (Ackerman and Okubo, 1993; Hurd and Stevens, 1997). Consequently,
species with high phenotypic plasticity may adopt a variety of morphologies to meet
environmental conditions without being developmentally committed to those conditions
should they change (Taylor and Hay, 1984). For example, in order to increase
photosynthetic ability, plants generally increase the size of the leaf surfaces thus
increasing the number of chloroplasts in contact with the environment (Rascio, 2002;
Ronzhina et al., 2004). Specifically, Elodea reduces the thickness of its leaves,
Ceratophyllum has a highly dissected morphology and, Vallisneria has few layers of
photosynthetic cells, due to the presence of inner aerenchyma lacunae, which can be
involved in supplying carbon dioxide (Rascio, 2002). Another plant feature that
increases photosynthesis is the ability of the plant to oscillate with the flow, thus causing
the fluid to become stirred (Koehl and Alberte, 1988). For example, algae with undulate
blades have been shown to oscillate with greater amplitude, but smooth blades oscillate

Chapter 3

112

with greater frequency (Koehl and Alberte, 1988). The greater amplitude would increase
local mixing, whereas the greater frequency would mean that less surface area is being
exposed to the flow, thus decreasing the drag forces.
The flexibility of the plant stem can also cause changes in flow patterns (e.g.,
canopies of flexible plants tend to have a logarithmic velocity profile whereas canopies of
more rigid plants do not have a logarithmic velocity profile), as was demonstrated in
canopies of Myriophyllum and Hydrilla (Shi and Hughes, 2002). It has also been shown
that morphology can affect the flow within the canopy. The flow within a canopy of
whorled and dissected leafed morphologies was reduced considerably more than the flow
in a linear-bladed leafed canopy, demonstrating that there was less interaction between
the flow and the linear-bladed morphology (Sand-Jensen and Mebus, 1996; Champion
and Tanner, 2000). This corresponds with the results seen in the angle of deflection data,
where the linear-bladed morphology {Vallisneria) changed the least with increasing
velocity. As such, it is likely that Vallisneria is better suited to higher flow environments
because it is less affected by the flow, whereas Ceratophyllum may be better suited to
slower flow environments because of its high level of interaction with the flow. It
appears that there is a strong relationship between plant morphology and fluid dynamics
that may have influenced the convergent evolution of these morphologies.
At the larger scale, an improved understanding of flow effects would aid in efforts
to restore streams and rivers, and maintain water quality as aquatic macrophytes act as
ecosystem engineers. This is especially important as many streams have been degraded
as the result of anthropogenic processes including: channelization, removal of riparian
vegetation, irrigation, pollutant discharges, and stormwater runoff (Poff et al., 1997).

Chapter 3

113

Restoration efforts often focus on improving physical habitat within the channel,
replanting streamside forests, and building stormwater retention structures. One of the
goals of these restoration practices is to create flow conditions that improve the “health”
of streams, as such it is important to link the physical aspects of the flow and the
biological aspects of the plant (Hart and Finelli, 1999), since freshwater macrophytes act
as ecosystem engineers in aquatic systems (Wright et al., 2004). It should be noted that
the range of velocities examined here are typical of within canopy conditions. If higher
velocities were obtainable, transport limitations (i.e., diffusional stress) would become
negligible, and there would be more compliance observed in the plants, less retention of
fluid, and ultimately dislodgment and loss of plant biomass.

Chapter 3

114

Literature cited
Ackerly, D.D., S.A. Dudley, S.E. Sultan, J. Schmitt, J.S. Coleman, C.R. Linder, D R.
Sandquist, M.A. Gerber, A.S. Evans, T.E. Dawson, andM.J. Lechowicz. 2000. The
evolution of plant ecophysiological traits: recent advances and future directions.
BioScience. 50(ll):979-995.
Ackerman, J.D., and T.M. Hoover. 2001. Measurement of local bed shear stress in
streams using a Preston-static tube. Lim nology an d O ceanography. 46(8): 2080-2087.
Ackerman, J.D., and A. Okubo. 1993. Reduced mixing in a marine macrophyte canopy.
Functional Ecology. 7:305-309.
Anderson, S.M., and A.C. Charters. 1982. A fluid dynamics study of seawater flow
through G elidium rudifrons. Lim nology an d O ceanography. 37(3):399-412.
Armstrong, S.L. 1989. The behaviour in flow of the morphologically variable seaweed
H edophyllum sessile (C.Ag.) Setchell. H ydrobiologia. 183:115-122.
Borchardt, M.A. 1994. Effects of flowing water on nitrogen- and phosphorus-limited
photosynthesis and optimum N:P ratios by S pirogyra flu via tilis (Charophyceae).
Journal o f Phycology. 30:418-430.
Bruchert, P., O. Speck, and H-C. Spatz. 2003. Oscillations of plants’ stems and their
damping: theory and experimentation. P h ilosoph ical Transactions o f the R oyal
Society o f London Series B -B iological Sciences. 358:1487-1492.
Carollo, E.G., V. Ferro, and D. Termini. 2002. Flow velocity measurements in vegetated
channels. Journal o f H ydraulic Engineering. 128(7):664-673.
Champion, P.D., and C.C. Tanner. 2000. Seasonality of macrophytes and interaction with
flow in a New Zealand lowland stream. H ydrobiologia. 441:1-12.
Cheer, A.Y.L, and M.A.R. Koehl. 1987. Paddles and rakes: fluid flow through bristled
appendages of small organisms. Journal o f Theoretical B iology. 129:17-39
Cook, C.D.K., B.J. Gut, E.M. Rix, J. Schneller, M. Seitz. 1974. W ater p la n ts o f the
w orld: a m anual f o r the identification o f the genera o f fre sh w a te r m acrophytes. Dr.

W. Junk b.v., publishers. The Hague.
Coutanceau, M., and J. Defaye. 1991. Circular cylinder wake configurations: A flow
visulization survey. A p p lied M echanics R eview s. 44(6):255-305.
Dawson, F.H. 1988. Water flow and vegetation of running waters, e.d Symoens, JJ
V egetation o f inland w aters. Kluwer academic publishers group. Pp.283-309.

Chapter 3

115

Dawson, F.H., and W.N. Robinson. 1984. Submerged macrophytes and the hydraulic
roughness of a lowland chalkstream. Verhandlungen d er In tem ation alen Vereinigung
fu r Lim nologie. 22(3):1944-1948.
Dodds, W.K., and B.J.F. Biggs. 2002. Water velocity attenuation by stream periphyton
and macrophytes in relation to growth form and architecture. Journal o f the N orth
A m erican B enthological Society. 21(1): 2-15.
Dorken, M.E., and S.C.H. Barrett. 2004. Phenotypic plasticity of vegetative and
reproductive traits in monoecious and dioecious populations in Sagittaria latifolia
(Alismataceae): a clonal aquatic plant. Journal o f Ecology. 92: 32-44.
Elliott, A.H. 2000. Settling of fine sediment in a channel with emergent vegetation.
Journal o f H ydraulic Engineering. 126(8):570-577.
Ennos, A.R. 1999. The aerodynamics and hydrodynamics of plants. The Journal o f
Experim ental B iology. 202:3281-3284.
French, T.D., and P.A. Chambers. 1996. Habitat partitioning in riverine macrophyte
communities. F resh w ater Biology. 36:509-520.
Ghisalberti, M., and H.M. Nepf. 2002. Mixing layers and coherent structures in vegetated
aquatic flows. Journal o f G eophysical Research. 107(C2): 3.1-3.11.
Givnish, T.J., 2002. Ecological constraints on the evolution of plasticity in plants.
E volutionary E cology. 16:213-242.
Goliber, T.E., and L.J. Feldman. 1990. Developmental analysis of leaf plasticity in the
heterophyllous aquatic plant H ippuris vulgaris. A m erican Journal o f Botany.
77(3):399-412.
Gutierrez, L.M., and C. Fernandez. 1992. Water motion and morphology in Chondrus
crispus (Rhodothyta). Journal o f Phycology. 28:156-162.
Kara, T., and S. Kato. 2004. Numerical simulation of thermal plumes in free space using
the standard K-e model. F ire Safety Journal. 39:105-129.
Hart, D.D., and C.M. Finelli. 1999. Physical-biological coupling in streams: The
pervasive effects of flow on benthic organisms. Annual R eview o f E cology an d
System atics. 30:363-395.
Hurd, C.L., P.J. Harrison, and L.D. Druehl. 1996. Effect of seawater velocity in
inorganic nitrogen uptake by morphologically distinct forms of M acrocystis
intgrifolia from wave-sheltered and exposed sites. M arine B iology. 126:205-214.

Chapter 3

116

Hurd, C.L. and C.L. Stevens. 1997. Flow visualization around single- and multiplelinear-bladed seaweeds with various morphologies. Journal o f P hycology. 33:360367.
Hurd, C.L, C.L Stevens, B.E. Laval, G.A. Lawrence, and P.J. Harrison. 1997.
Visualization of seawater flow around morphologically distinct forms of the giant
kelp M acrocystis integrifolia from wave-sheltered and exposed sites. Lim nology and
O ceanography. 42(1): 156-163.
Irwin, S., and J. Davenport. 2002. Hyperoxic boundary layers inhabitated by the
epiphytic meiophauna of Fucus serratus. M arine E cology P ro g ress Series. 244:7379.
Jarvela, J. 2002. Flow resistance of flexible and stiff vegetation: a flume study with
natural plants. Journal o f H ydrology. 269:44-54.
Johnson, A.S. 2001. Drag, drafting, and mechanical interactions in canopies of the red
alga Chondrus crispus. B iological Bulletin. 201:126-135.
Jumars, P. A., J.E. Eckman, and E. Koch. 2001. M acroscopic anim als an d p la n ts in
benthic flo w s. In. The benthic boundary layer: Transport processes and
biogeochemistry. B.P. Boudreau ed. Oxford University Press. Pp. 320-347.
Kiya, M., O. Mochizuki, and H. Ishikawa. 2000. Challenging issues in separated and
complex turbulent
turbuleni flows. P roceedin gs o f the 70^ International Conference on L azer
Techniques. 1-13.
Koch, E.W. 1993. The effect of water flow on photosynthetic processes of the alga Ulva
lactuca L. H ydrobiologia. 260/261:457-462.
Koch, E.W. 1994. Hydrodynamics, diffusion-boundary layers and photosynthesis of the
seagrass Thalassia testudinum and C ym odocea nodosa. M arine Biology. 119:767-776.
Koehl, M.A.R. 1996. When does morphology matter? AnnwaZ R eview o f E cology and
System atics. 27:501-542.
Koehl, M.A.R., and R.S. Alberte. 1988. Flow, flapping, and photosynthesis of
N ereocystis luetkeana: a functional comparison of undulate and flat blade
morphologies. M arine B iology. 99:435-444.
Kundu, P.K. 1990. F luid M echanics. Academic Press, San Diego, Toronto. 638 p.
Leonard, L.A., and M.E. Luther. 1995. Flow hydrodynamics in tidal marsh canopies.
Lim nology an d O ceanography. 40(8): 1474-1484.

Chapter 3

117

Losee, R.F., and R.G. Wetzel. 1993. Littoral flow rates within and around submersed
macrophyte communities. F reshw ater Ecology. 29:7-17.
Macdonald, R.W., R.K. Strom, and P R. Slawson. 2002. Water flume study of the
enhancement of buoyant rise in pairs of merging plumes. A tm osph eric Environment.
36:4603-4615.
Mercier, G.P., and Y. Jaluria. 1999. Fire-induced flow of smoke and hot gases in open
vertical enclosures. Experim ental Therm al an d F luid Science. 19:77-84.
Nepf, H.M. 1999. Drag, turbulence, and diffusion in flow through emergent vegetation.
W ater R esources Research. 35(2):479-489.
Nepf, H.M., and E.W. Koch. 1999. Vertical secondary flows in submersed plant-like
arrays. Lim nology a n d O ceanography. 44(4): 1072-1080.
Norberg, C. 2003. Fluctuating lift on a circular cylinder: review and new measurements.
Journal o f Fluids an d Structures. 17:57-96.
Norton, T.A., A.C. Mathieson, and M. Neushul. 1982. A review of some aspects of form
and function in seaweeds. B otanica M arina. XXV: 501-510.
Nowell, A R M., and P.A. Jumars. 1984. Flow environments of aquatic benthos. Annual
R eview o f E cology a n d System atics. 15:303-328
Palmer, M R., H.M. Nepf, T.J.R. Pettersson, and J.D. Ackerman. 2004. Observations of
particle capture on a cylindrical collector: implications for particle accumulation and
removal in aquatic systems. L im nology an d O ceanography. 49(l):76-85.
Poff, N.L., J.D. Allan, M.B. Bain, J R. Kar, K.L. Prestgaard, B.D. Ritcher, R E. Sparks,
and J.C. Stromberg. 1997. The natural flow regime. B ioscience. 47(ll):769-782.
Rascio, N. 2002. The underwater life of secondary aquatic plants: Some problems and
solutions. C ritical R eview s in P lan t Sciences. 21(4):401-427.
Riemer, D.N. 1993. Introduction to F reshw ater Vegetation. Krieger Publishing
Company, Malabar, Florida.
Ronzhina, D.A., G.F. Nekrasova, and V.l. P ’yankov. 2004. Comparative eharacterization
of the pigment complex in emergent, floating, and submerged leaves of hydrophytes.
Russian Journal o f P lan t P hysiology. 51(l):21-27.
Sand-Jensen, K., and J R. Mebus. 1996. Fine-scale patterns of water velocity within
macrophyte patches in streams. Oikos. 76:169-180.

Chapter 3

118

Sand-Jensen, K., and O. Pedersen. 1999. Velocity gradients and turbulence around
macrophyte stand in streams. F reshw ater B iology. 42:315-328.
Santamaria, L. 2002. Why are most aquatic plants widely distributed? Dispersal, clonal
growth and small-scale heterogeneity in a stressful environment. A cta O ecologica.
23:137-154.
Santamaria, L., J. Figuerola, J.J. Pilon, M. Mjelde, A.J. Green, T. De Boer, R.A. King,
and R.J. Gomall. 2003. Plant performance across latitude: the role of plasticity and
local adaptation in an aquatic plant. Ecology. 84(9):2454-2461.
Schlichting, C.D. 1986. The evolution of phenotypic plasticity in plants. Annual R eview
o f E cology an d System atics. 17:667-693.
Schutten, J., and A.J. Davy. 2000. Predicting the hydraulic forces on submerged
macrophytes from current velocity, biomass and morphology. O ecologia. 123:445452.
Schwinge, J., D.E., Wiley, and D.F. Fletcher. 2002. Simulation of the flow around spacer
filaments between narrow channel walls. 1. Hydrodynamics. Industrial an d
Engineering C hem istry Research. 41:2977-2987.

Sheath, R.G., and J.A. Hambrook. 1988. Mechanical adaptations to flow in freshwater
red algae. Journal o f P hycology. 24:107-111.
Shi, Z. and J.M.R. Hughes. 2002. Laboratory flume studies of microflow enviroments of
aquatic plants. H ydrological P rocesses. 16:3279-3289.
Stevens, C.L., and C.L. Hurd. 1997. Boundary-layers around linear-bladed aqutic
macrophytes. H ydrobiologia. 346:119-128.
Taylor, P.R., and M.E. Hay. 1984. Functional morphology of intertidal seaweeds:
adaptive significance of aggregate vs. solitary forms. M arine E cology P ro g ress Series.
18:295-302.
Vogel, S. 1987. Flow-assisted mantle cavity refilling in jetting squid. B iological Bulletin.
172:61-68.
Vogel, S. 1994. Life in M oving Fluids the P h ysical B iology o f F low 2”^ ed. Princeton
University Press, Princeton, N.J. 467 p.
Wiegleb, G. 1984. A study of habitat conditions of the macrophytic vegetation in selected
river systems in western lower saxony (Federal Republic of Germany). A qu atic
Botany. 18:313-352.

Chapter 3

119

Williamson, C.H.K. 1996. Vortex dynamics in the cylinder wake. Annual Review o f Fluid
Mechanics. 28:477-539.
Wright, J.P., W.S.C. Gurney, and C.G. Jones. 2004. Patch dynamics in a landscape
modifies by ecosystem engineers. Oikos. 105:336-348.
Wuest, A., and A. Lorke. 2003. Small scale hydrodynamics in lakes. Annual review o f
fluid mechanics. 35:373-412.

120

Box 3.1: Matlab commands for the image analysis.

1) Import file
I=imread('c:\julianne\photoshop\elodea\4inc6ta.png');
2) Edit image
I=rgb2gray(I);
Convert to gray scale
I=imadjust(I,[0 0.2],[0 1]);
Adjusts the intensity of the image.
Pixval
Gives pixel value at mouse click.
I=imsubtract(I,69);
Subtracts background value (use value given by above
command)
3) Crop image
I=imcrop(I);
Use mouse to select area based on lines drawn on the flume
I=imresize(I, [600 900], 'bicubic') ;
To make image larger and thus easier to see; the matrix
dimensions change depending on domain.
4) Extract pixel values
[n x]=imhist(I,1 0 0 );
Groups the pixel values into 100 different bins
n
Gives the number of pixels in each bin

121

Table 3.1: Shear velocity (u*) determined for each of the velocities used for the
U (cms‘^)
1.3
2.0
5.0
8.4
11.0

u* (cms'^)
0.12
0.24
0.42
0.33
0.35

122

Table 3.2: Pump settings for isokinetic release of dye at each water velocity.
Water Velocity
(ems'^ ± St.dev)
1.3 ±0.1
2.0 ±0.5
5.0 ±0.3
8.4 ±0.5
11.0 ±0.9

Pump Speed

Increment

Slow
Slow
Fast
Slow
Slow

0
3
10
2
6

Internal Tubing
Diameter
3/32” = 0.21 cm
3/32” = 0.21 cm
3/32” = 0.21 cm
3/16” = 0.42 cm
3/16” = 0.42 cm

123

Table 3.3; ANCOVA results for the area of dye coverage and the concentration of dye for
(a) the empty test section, (b) the circular cylinder, (c) Vallisneria, (d) Elodea, and (e)
Ceratophyllum. (Area = area of dye coverage; Cone. = concentration of dye; velocity =
average flume vélocités; side/top = side and top perspectives; before/after = upstream and
(a)
Effect
Intercept
velocity
side/top
before/after _
side/top'before/after
Error
Total

Univariate Resuits for Each DV (Empty Test Section)
Sigma-restricted parameterization
Effective hypothesis decomposition
conc.
conc.
Degr. of I area j area j area
area ) conc. i conc.
Freedom I S S
1 MS 1
F
p
1 SS
i MS
F
P
ll 15.896351 15.89635 20.600621 0.0003921 5.9861951 5.9861951 80.562241 0.000000
...... 1| 0.02138 0.02138 0.02771 0.870021 1.9889601 1.9889601 26.767431 0.000113
11 2.681131 2.68113! 3.474561 0.082011 0X1002791 0.0002791 0.00375)0.951949
Î 3Î 75762 31.75762 41.15577 0.000012 3^ 382251 0.5362251 7.216521 0.01 # 1 5
1 3.63(g2l 3.630221 4 . 7 0 - ; . ' , Ii (i-'K ea 0.028648 0.028648! 0.385541 0.54M 72
i
1'l ï ï l 4 5 7 Ï a 0 7 4 3 d 5 l
19 11.57467 0 77164
19| 48 66502
1 3.668690}

Univariate Results for Each DV (Cylinder)
Sigma-restricted parameterization
Effective hypothesis decomposition______
area
area
area
Degr. of area
Freedom- S3
MS
P
22 059. 22.05981 22.2123 0.00027
1
5 . ^ 5.43651 5.474110.03354
J
0.4341; 0.4345ii 0.43751! 0.51833
lide^op
1 95.6831; 95.6839 96.3455 0.00000
1
0.3294 0.3294.1 0.3317: 0.57318
side/top;before/ajter
14.8971^ o"993ll
Error
116.781:
T otaf""................. ^

(b)

(c)
Effect
velocity

side/tqp,_______
.before/after...........
side/top’before/after

Error...................
Total

(d)
Effect
Intercept
velocity^
side/top

More/attex_____
slde/top*befora(after„
Error.............. _
Total

(e)
Effect
lnt<

veiogiy..

side./tp2_...
before-after
side/top'before/after
Error
Total

conc.
S3
1.97869
0.437851
0.01650
0.03349;
0.00042;
0.310231
0.79851:

1.97869:
0.43785t
0.01650;
0.03349
0.00042
0.02068:

95.6720;
21.1706:
0 7979
1 6196.
0.0207(1

0.00000
0.00034
0.38580
0.222511
0.88752

Univariate Resuits for Eacfi DV { Vallisneria)
Sigma-restricted parameterization
Effective fiypothesis decomposition
conc,
conc
area
area
conc.
Degr. of j area
area
conc.
F
MS
F_________e_
.
SS
MS
Freedom
SS
__e_
25.68525! 25.685251 60.88711 O.OOOOO1I 2.074531 2.074531! 37.30564! 0.000020
2.076881 2 . 0 7 6 ^ 4.!
84: 0.199396 0.199396! 3.58567 0.077732
0.36396! 0.36396 0.8628 0.36767: 0.002246 0.002246! 0.04038 0.843438
81.817581 81.817M 193 9494 OOOOOOC 0 . 0 1 2 ^ 0 0.012980! 0.23342 0.635978
0.455471 0.455471 1.07971 0.31522: 0.003279 0.003279! 0.058971 0.811423
0.834Ï36 0.055609!_________ I
6.32775! 0 42185 ________ I
1.052036
91.04164!

19

Univariate Results for Each DV ( Elodea)
Sigma-restricted parameterization
Effective hypothesis decomposition
Degr. of
area
area
area
Freedom !
SS
MS
F
40.6380 40.637981 111.6627 o .o o o o o d
0.1589!
0.158881 0,4366 0.51880
0.4885
0.488551 1.3424 0.26472
^92.4244192.42442 253.9585 0.00000
^ 1.77681
1.776811 4 . 6 ^ 0.04310
5.4590 0.363941
19 100.30771

conc.
conc. ! conc.
F
SS
MS
2.9495231 2.949523! 120.1269! 0.000000
0.6493321 0.6493321 26.44571 0.000120
0.029748! 0.029748!
1.21161 0.288375
0 028255 0 028255
1 1508 0.300344
0.0789951 0.078995!
3.21731 0.093045
0.368301! g ^ ^ 5 3 !
1.1546311

Univariate Results for Each DV ( Ceratophylluni)
Sigma-restricted parameterization
Effective hypothesis decomposition
area
of
area
are
Freedom
MS :
S3
■13.71454: 43.714541 109.4433
1.35136
1.351361 3.3833 0.08573Ï
1.69315
1.69315 4.2390
65.794651 65.7946S 164.^ 8 0 . 00000:
7.873771
1.87377 4.% 12
5.99140 0 39943
76.70433

conc.
conc.
conc.
F
SS
MS
__
3.4922961 3.492296! 32.721281 0.000041
0.446894 0 446894 4.18720 0.058675
0,152006! 0.152006! 1 .4 2 4 2 ^0 ,2 5 1 2 4 3
0.005253! 0 . 0 & 2 ^ 0.0492lT g827427
0.033108! 0.033108 0.31021 0 585767
1.600929! 0.1067297........
l"
2.238190’

gÔ468&

124

Table 3.4; Post hoc comparisons for the area of dye coverage (Table 3) in the different
domains surrounding different test subjects: (a) empty test section, (b) circular cylinder, (c)
Vallisneria, (d) Elodea, and (e) Ceratophyllum. Significant results indicated in bold.
(a)

Empty flume side; upstream

top; upstream

side; downstream

top; downstream

side; upstream
top; upstream

0.81

side; downstream

< 0.001

< 0.001

top; downstream

< 0.001

0.002

0.003

top; upstream

side; downstream

(b)

Cylinder side; upstream

top; downstream

side; upstream
top; upstream

0.94

side; downstream

< 0.001

< 0.001

top; downstream

< 0.001

< 0.001

0.28

side; upstream

top; upstream

side; downstream

(c)

Vallisneria

top; downstream

side; upstream
top; upstream

0.95

side; downstream

< 0.001

< 0.001

top; downstream

< 0.001

< 0.001

0.26

side; upstream

top; upstream

side; downstream

(d)

Elodea

top; downstream

side; upstream

(e)

top; upstream

0.58

side; downstream

< 0.001

< 0.001

top; downstream

< 0.001

< 0.001

0.078

Ceratophyllum

side; upstream

top; upstream

side; downstream

side; upstream
top; upstream

0.95

side; downstream

< 0.001

< 0.001

top; downstream

< 0.001

< 0.001

0.021

top; downstream

125

Table 3.5: Post hoc comparisons for the concentration of dye (Table 3) in the different
domains surrounding different test subjects: (a) empty test section, (b) circular cylinder, (c)
Vallisneria, (d) Elodea, and (e) Ceratophyllum. Significant results indicated in bold.
(a)

Empty flume

side; upstream

top; upstream

side; downstream

top; downstream

side; upstream

(b)

top; upstream

0.59

side; downstream

0.011

0.042

top; downstream

0.034

0.11

0.66

side; upstream

top; upstream

side; downstream

Cylinder

top; downstream

side; upstream
top; upstream

0.76

side; downstream

0.56

0.37

top; downstream

0.88

0.67

0.64

side; upstream

top; upstream

side; downstream

(c)

Vallisneria

top; downstream

side; upstream
top; upstream

0.76

side; downstream

0.62

0.84

top; downstream

0.64

0.87

0.97

side; upstream

top; upstream

side; downstream

(d)

Elodea

top; downstream

side; upstream

(e)

top; upstream

0.76

side; downstream

0.75

0.99

top; downstream

0.33

0.20

0.20

Ceratophyllum

side; upstream

top; upstream

side; downstream

side; upstream
top; upstream

0.55

side; downstream

0.46

0.19

top; downstream

0.36

0.75

0.10

top; downstream

126

Table 3.6: ANCOVA results for (a) the area covered in dye and (b) the concentration of
dye in that area for the different domains. (Velocity = covariate; Plant = empty test section,
circular cylinder, Vallisneria, Elodea, and Ceratophyllum', Side/Top = perspectives; and
Before/After = upstream and downstream).
(a)

Effect
Intercept
velocity
plant
side/top
before/after
plant*side/top
plant*before/after
slde/top*before/after
plant*slde/top*before/after
Error
Total

(b)

Effect
Intercept
velocity
piant
side/top
before/after
piant’ side/top
p ia nt'b efo re/a fte r
side/top* before/after
pla nt'side /to p ’ before/after
Error
Totai

Univariate Resuits for Each DV (Spreadsheets)
Sigma-restricted parameterization
Effective hypothesis decomposition
area
area
Degr. of
area
F
Freedom
SS
MS
220.1792
1
142.9880
142.98801
3.0657
1.9909
1
1.9909
5 6756
4
14.7433
3.1,
2.7284
4.2013
1
2.7284
348.1224
355.9598
1
355.9598
0.7332
1.1291
4
2.9329
4.4341
2.8796
4
1 1.51^^ __
6.2537
1
4.0612
4.0612|
1 5416
4
4.0045
1.0011
0.6494
51.3039]
79
449.24341
99

area
P
0.000000
0.083840
0.000456
0.043711
0.000000
0.348892
0.002764
0.014463
0.198277

Univariate Resuits for Each DV (Spreadsheets)
Sigm a-restricted param eterization
Effective hypothesis decom position
concentration
concentration
concentration
concentration
F
SS
MS
P
261.623C
0.000000
15.77032
15.77032
3.18859
0.44452
0.11398
0.11516
0.08680
0.50106
0.01323
0.13123
4.76202
9.35658

3.18859
0.11113
0.11398
0.11516
0.02170
0.12526
0.01323
0.03281
0.06028

52.8973
1.8436
1.8909
1.9104

0.000000
0.128760

0.3600
2.0781
0.2195

0.836338
0.091504
0.640741

0.5443

0.703693

0.172993
0.170817

127

Table 3.7: Post hoc comparisons between the subjects for the area covered in dye (Table 6 )
in the different domains; (a) side upstream, (b) top upstream, (c) side downstream, and (d)
top downstream. Significant results indicated in bold.
Cylinder Vallisneria

(a)

Elodea

Ceratophyllum

Empty flume

Cylinder
Vallisneria

0.98

Elodea

0.85

0.83

Ceratophyllum

0.91

0.89

0.94

Empty flume

0.75

0.73

0.89

Cylinder Vallisneria

(b)

Elodea

0.83
Ceratophyllum Empty flume

Cylinder
Vallisneria

0.97

Elodea

0 .6 6

0 .6 8

Ceratophyllum

0.98

0.99

0.67

Empty flume

0.99

0.96

0.65

Cylinder Vallisneria

(c)

Elodea

0.97
Ceratophyllum Empty flume

Cylinder
Vallisneria

0.090

Elodea

0.74

0.043

Ceratophyllum

0.38

0.41

0.23

Empty flume

0.007

0.28

0 .0 0 2

Cylinder Vallisneria

(d)

Elodea

0.059
Ceratophyllum Empty flume

Cylinder
Vallisneria

0.63

Elodea

0.72

0.39

Ceratophyllum

0.036

0 .0 1 0

0.080

Empty flume < 0 .0 0 1

< 0 .0 0 1

< 0 .0 0 1

0.009

128

Table 3.8: Post hoc comparisons between the subjects for the concentration of dye in the
plume (Table 6) in the different domains; (a) side upstream, (b) top upstream, (c) side
downstream, and (d) top downstream. Significant results indicated in bold.

(a)

Cylinder

Vallisneria

Elodea

Ceratophyllum

Empty flume

Cylinder
Vallisneria

0.90

Elodea

0.77

0.86

Ceratophyllum

0.38

0.46

0.57

Empty flume

0.73

0.64

0.52

Cylinder

Vallisneria

Elodea

(b)

0 .2 2

Ceratophyllum Empty flume

Cylinder
Vallisneria

0.91

Elodea

0.75

0.66

Ceratophyllum

0.25

0.30

0.14

Empty flume

0.90

0.81

0.84

Cylinder

Vallisneria

Elodea

(c)

0 .2 0

Ceratophyllum Empty flume

Cylinder
Vallisneria

0.23

Elodea

0.58

0.52

Ceratophyllum

0.47

0.64

0.87

Empty flume

0.005

0.11

0.025

Cylinder

Vallisneria

Elodea

(d)

0.038
Ceratophyllum Empty flume

Cylinder
Vallisneria

0.46

Elodea

0.15

0.49

Ceratophyllum

0.055

0.23

0.61

Empty flume

0.052

0.23

0.60

0.99

129

Table 3.9: ANCOVA results for velocity in the (a) streamwise component (u) and (b)
vertical component (w) for the test subjects. (Velocity = different test velocities (covariate);
Plant = empty test section, circular cylinder, Vallisneria, Elodea, and Ceratophyllum-, and
Before/After = upstream and downstream).
(a)

Effect

Univariate Results fo r Each DV (stream wise velocity (u))
Sigm a-restricted param eterization
Effective hypothesis decom position
Degr. of
Freedom
1
1
4
1

Intercept
velocity
plant
before/after
plant*before/after

4

vel (x)
88
:
72.107
3819.222
2.998
4.424!
3.877

Error

289

245.5331

Total

299

4079.8991

,( b )

vel (x)
MS
72.107
3819.222
0.750
4.424

F

vel (z)
SS
0.07739

vel (z)
MS
0.077394

velocity

0.30981

0.309814

plant

1.00565

0.251412

before/after
p la nt'b efo re/a fte r

1.28823
1.03662

Error
Total

14.80655
18.85226)

1.288225
0.259155
0.051234

1

84.8731
4495.343
0.882
5 207

0.969
0.850

1.141
1

U nivariate Results for Each DV (vertical velocity (w))
Sigm a-restricted param eterization
Effective hypothesis decom position

Effect
Intercept

vel (x)

vel (x)

vel (z)
F

vel (z)

1.51060

P
0,220048

6.04708
4.90716

0.014515
0.000766

25.14408
5.05828

O.OOUUdl
0.000592
1

P
0.000000
0.000000
0.474875
0.023220
0.337392

130

Table 3.10: Post hoc comparisons for velocity (Table 8 a,b) in the (a) streamwise
component (u), and (b) vertical component (w) for the different test subjects. (Cell No.:l =
upstream of the cylinder; 2 = downstream of the cylinder; 3 = upstream of Elodea; 4 =
downstream of Elodea; 5 = upstream of Ceratophyllum; 6 = downstream of Ceratophyllum;
7 = upstream of Vallisneria; 8 = downstream of Vallisneria; 9 = upstream in the empty test
section; and 1 0 = downstream in the empty test section).
(a)

LSD test; variable vel (x) (stream w ise velocity (u))
Probabilities for P ost Hoc T ests
Error: B etw een MS = .84960. df = 2 8 9 .0 0
{10}
{6}
{8}
{7}
{2}
{3}
{4} 1 {5}
{9}
6 .0 6 6 0 6 5 5 9 3 5.9762: 6 .3 3 1 3 6.1044: 6 .4 3 8 0 5 .8 5 5 3 ' 6 .4 0 4 7 6 .4 5 8 4
1
0.81785 0 20241 0.57785 0.54885 0.92785 0.35985 0.31957 0.41395 0.23205
2 0.81785
0.07365 0.64442: 0.33505 0 .8 4 3 3 0 0.17685 0.27921 0.21881 0.04434
1 0.20241: 0.07365
0.034651 0.49865 0 09895 0.71873 0 .01092 0.646151 0.71375
0.19731 0.50987 0.09395) 0.53435 0.12007 0 01364
2 0 57785 0.64442 0.03465!
1 0.54885 0.335051 0.49868 0.19731
0 40965 0.75153 0.08432 0 8 2 7 6 7 0.64403
0.22582 0.20088 0 27553 0.06952
2 0.92785 0.8433C: 0.09895 0.50987 0 40965,
1 0.35985 0.17685 0.71873 0.09395 0.75153 0 22582
0.03483 0.92117 0.94074
0.04654 0.0021 C
0 31957 0.27921 0.01092 0.53435 0.08432 0.20088 0.03483
1 0.41395 0.21881: 0.64615 0.12007 0.82767 0.27553: 0.921171 0.04654
0.84498
2 0.23205 0.044341 0.71375 0.01364: 0.64403 0.06952 0.94074 0.0021 Cl 0.84498

plant before/aftei
Cell No
1
2
3
4
5

1|
1:
2i:
2i
51

6
7
8
9
10

(b)

51
8
8
g:
9|

LSD test; variable vel (z) (vertical velocity (w))
Probabilities for P o st Hoc T ests
Error: B etw een MS = .05123, df = 2 8 9 .0 0
plant before/aftei

Cell No

1
2
3
4
5

6
7

8
9

10

{1}
6 .1 2 9 3

{ 10 }
{2}
{3}
{8 }
{4} I
{5}
{61 I {7}
{9}
-.1 1 4 0 -.1202
-.3 2 2 0 .19333
.12181 -.0587 -.1 1 7 6 ' -.0 2 7 3 -.0871
O.OOOOC 0.0168G 0 0855C 0.51925 0.0974c: 0.79025 0.22657 0 .18963 0.08974
o.ooooc
o.oobod 6 00002 0.00011 0.00002 0.00001 O.OOOOC: 0.00225: 0.00003
10 ooooc 0.0025C O.OOOOC 0.00801 0 0 0 0 0 4 0.00024 O.OOOOC
0,01686 O.OOOOC
0.35045 0.92955 0.16273 0.46813 0 .90832 0.97401
0.0855C 0 . 0000^ o.ooooc
0 38355 0.70485 0.6737C 0 .50372 0 .36245
0 .51925 000011 0.0025C 0 3504C
0.18225 0.52397 0.958011 0 .9 5 5 4 7
0.0974C oooooc o.ooooc 0 92955 0.38355
0.37645 0 .29524 0.16974
0.7902Ê 0.00001 0.00801 0 16273 0.70485 0.18225:
0.69055 0.48831
0.22657 o.ooooc 0 .0 0 0 0 4 0 46813 0 6737C 0.52397: 0.37645:
0.9266C
0.1896S 0.002251 0 .0 0 0 2 4 0 90832: 0 50372 0.95801 0.29524 0 69055
0 .08974 o.ooooc! o.ooooq 0 97401 0.362451 0.95547 0.16974 0.48831 0.9266C

(1)
-.0 0 5 3

131

Table 3.11: ANCOVA results (a) and post hoc comparison (b) for examining the angle of
deflection and the different test subjects. (Velocity = covariate; and Plant = circular
cylinder, Vallisneria, Elodea, and Ceratophyllum). Significant differences indicated in
bold.
(a)

Effect
Intercept
velocity
plant
Error
Total

Univariate Results for Eacfi DV (angle)
Sigma-restricted parameterization
Effective fiypotfiesis decomposition
Degr. of ' angle ■ angle
angle
Freedom '
SS
MS
F

I angle
p

1 3906.17 3906.167 36.64167 0.000008
1 1761.57 1761.5751 16.52439 _g,000660
3 11849.58 3949.859 37.05153 0.000000
Ï9 2025.49 106.604
23 15636.64
i:

(b)

Cylinder

Vallisneria

Elodea

Cylinder
Vallisneria

< 0.001

Elodea

0.008

< 0.001

Ceratophyllum

< 0.001

0.023

< 0.001

Ceratophyllum

132

Table 3.12: ANCOVA results (a) and post hoc comparisions (b) for the frequency of
oscillations between the different test subjects. (Velocity = covariate; and Plant = circular
cylinder, Vallisneria, Elodea, and Ceratophyllum). Significant differences indicated in
bold.
(a)

Effect
Intercept
velocity
plant
Error
Total

Univariate Results for Eacfi DV (oscillations)
SIgma-restrlcted parameterization
Effective hypothesis decomposition
Degr. of oscillations per ; oscillations per oscillations per
Freedom
second
second
second
F
SS
MS
1
1.27783
0.0309081
0.030908
1
29.82397
0.721379
0.721379
0.Ô32280
0.177-127
3
7.33535
0.459570
0.024188
19
23
1.713229
1

(b)

Cylinder

Vallisneria

Elodea

Cylinder

Vallisneria

< 0.001

Elodea

0.001

0.73

Ceratophyllum

0.002

0.61

0.87

Ceratophyllum

133

Black light

Microscope
light —

Flow straighteners
Dye injector

15 cm

Outlet

Inlet
1 2 0 cm

Figure 3.1: Lighting and dye injector set up for the photos (flume dimensions: 19.2 cm (width) x 24.7 cm (height) x 170 cm (length)
with water depth ranging from 9.7-13.3 cm from the lowest (1.27 cms"') to the highest (11.0 cms'*) test velocities)

134

0.05
0.05
V = 0 .00020 ^®
R2-0PP14

(a )
0.04

0.04
0.03

0.03

0.02

0.02

0.01
0

y

0.01
1

1

1

0.05

0.1

0.15

0.05

0.2

0.05

0.1

0.15

0.2

0.15

0.2

0.05

90.439X

y = 0.0001e-

= 0 .9846

0.04

S

0.04

0.03

0.03

0.02

0.02

0.01

0.01

0.05

0.1

0

0.2

0.15

0

0.05

0.1

0.05
0.04

y = 1 .1 7 5 5 x - 0.1223
R® = 0.9894

0.03

0.02
0.01

0.05

0.1

0.15

0.2

Velocity ms'
Figure 3.2: Velocity profiles in the empty test section at: (a) 0.013 m s'\ (b) 0.020
m s'\ (c) 0.050 ms , (d) 0.084 ms‘\ and (e) 0.11 m s'\ Each point represents the
mean of 120 s of data samples taken at 25 Hz using an acoustic doppler
velicometer (ADV).

135

M

J

ï

î

>

V
9.7 c m 13.3 cm
(range of
water depth
with
increasing
velocity:
1.3 cm/s
11,0 cm/s)

Flume wall
Plexiglas plate

Syringe

Cylinder

f tape
Black
covering Plasticine

3cm

False bottom

Flume bottom
<------19.2 cm

Figure 3.3: Diagram of the positioning device attached to the side of the flume
with the cylinder in place for the top perspective.

136

S e ria l

Dllulio
XXX

Area where
pixels were
sampled

Black tape on up
and downstream
sides of the
cuvettes

Figure 3.4: Cuvette set up for creating the concentration curve.

137

(a)

200

180
160
140

q)

3 120
100

80
60
40
20 t
0
0

20

40

60

80

100

concentration (%)

C o n cen tratio n %

Figure 3.5: Original concentration curve (pixel value (i.e., intensity) vs. dye
concentration) for the (a) side and (b) top perspectives. The dotted line is at the 10%
concentration point, where the graph was divided into a linear and non-linear
components.

120

138

2 .2

200

2 .0

190
® 180
± 170

CL

Q. 160
150
140
•2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0

0

20

log % concentration

40

60

80

100

120

100

120

% concentration
100

® 1 .6

I

85
80

§ 1.4
0 1 .2
1 .0

CL

2 0 .8

-2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5
log % concentration

70
65
60
0

20

40

60

80

% concentration

Figure 3.6: Pixel intensity vs. dye concentration curves, (a) log (pixel value) versus log
(dye concentration) in the side perspective for 0.01-5% dye, (log pixel value = (0.30 ±
0.02) log %concentration + (1.86 ± 0.02); r^ = 0.98; p = 0.001). (b) pixel value versus
dye concentration in the side perspective for 10-100% dye (pixel value = (0.44 ± 0.03)
%concentration (143 ± 2); r^ = 0.98; p < 0.001). (c) log (pixel value) versus log (dye
concentration) in the top perspective for 0.01-5% dye, (log pixel value = (0.36 ± 0.02) log
%concentration (1.50 ± 0.02); r^ = 0.98; p < 0.001). (d) pixel value versus dye
concentration in the top perspective for 10-100% dye (pixel value = (0.25 ± 0.02)
%concentration -t- (63.5 ± 0.9); r^ = 0.99; p < 0.001).

139

■

Figure 3.7: Mock photo used to test image analysis method in MATLAB.

140

1

12

14 15

Circular
Cylinder

Figure 3.8: Grid used to take ADV measurements in the test section (120 cm downstream
of the flow straighteners). The grid was drawn on the top of the flume wall in the
downstream direction, and on a sliding piece of plexiglass in the cross stream direction.
The centre of the probe was aligned with the centre of both lines.

141

Figure 3.9: Flow visualization in the empty test section using fluoroscein dye in the side
(left) and the top (right) perspectives. Velocities increase from top to bottom panels [(a)
and (b) are 1.3 cms' , (c) and (d) are 2.0 cm s'\ (e) and (f) are 5.0 cm s'\ (g) and (h) are
8.4 cms ', and (i) and (j) are 11.0 cms ']. Flow is from left to right. Top and side photos
were taken sequentually.

142

Side

Top
Chamber Reynolds number (x 10^)

0.0 1.1

2.2 3.3 4.4 5.5

6 .6

0.0 1.1

2.2 3.3 4.4 5.5

6.6

(a) y = (0.009± 0.01)x + 0.27±0.08

(b) y = (0.009±0.01)X + 0.2710.08

= 0.14, p = 0.52

= 0.14, p = 0.52

y = (-0.1± 0.2)x + 4±I
= 0.09, p = 0.61

y = (0 .1 2 1 0 .0 9 )x + 1.410.6
f = 0.27, p = 0.26

y = (-0.01± 0.03)x + 0.4+0.2
,.2 _ n no „ _ n AO

y = (-0 .0 7 1 0 .0 2 )x + 0.810.1

y = (-0 .1 5 ± 0 .0 3 )x + 1.510.2
= 0.90, p = 0.012

y = (-0 .1 0 1 0 .0 1 )x + 1.210.1
? = 0.93, p = 0.007

H. 1 0

I

2 .0

(e )

= 0.80, p = 0.039

(D 1 .0

§

A 0.5

"o
§

0 .0

1

2 .0

g

Iu
Q
0.5
0 .0
0

2

4

6

8

10

12

0

2

4

6

8

10

12

Velocity (cms'^)
Figure 3.10: Regression of area of dye coverage (a-d) and concentration of dye (e-h)
versus velocity or Reynolds number (based on the hydraulic diameter) in the different
domains for the empty test section: (a) and (e) upstream in the side perspective, (b) and
(f) upstream in the top perspective, (c) and (g) downstream in the side perspective, and
(d) and (h) downstream in the top perspective. Solid line is the regression and dotted
lines are the 95% confidence limits of the mean.

143

Figure 3.11: Flow visualization around a 0.7 cm diameter circular cylinder using
fluoroscein dye in the side (left) and the top (right) perspectives. Velocities increase
from top to bottom panels [(a) and (b) are 1.3 cm s'\ (c) and (d) are 2.0 cm s'\ (e) and (f)
are 5.0 cm s'\ (g) and (h) are 8.4 cms'*, and (i) and (j) are 11.0 cms'*]. Flow is from left
to right. Top and side photos were taken sequentually.

144

Side
0.0

1.5

Top

Reynolds number (x 10 )
3.0

4.5

6.0

7.5

0.0

1.5
(b )

(a) y = (0.01 i±o.oo 6)x + 0.43+0.04

3.0

4.5

6.0

7.5

y = (0 .003± 0.002)x + 0.4+0.1
= 0.36, p = 0.28

^ —c\ An „ _ n 10

eg
s

y = (0.5± 0.1)x + 2.4±0.7
= 0.88, p = 0 .0 1 ^

II

y = (0.05± 0.05)x + 4.3±0.3
^ = 0.22, p = 0.41

6

C/3

Q

4
2
0
2 .0
(e )

y = (-0.05± 0.02)x + 0 .6 ± 0 .1
= 0.76, p = 0.053

y = (- 0 .0 3 ± 0 .0 2 )x + 0.6±0.2
= 0.31, p = 0.32

(■p) y = (-0.05± 0.01)x + 0.54± 0.06

y = (-0.03± 0.02)x + 0.5+0.1
= 0.47, p = 0.19

0 .0

2 .0

°

_ n CO - n nij

0.5
0 .0
0

2

4

6

8

10

12

0

2

4

6

8

10

12

Velocity (cms"')
Figure 3.12: Regression of area of dye coverage (a-d) and concentration of dye (e-h)
versus velocity or Reynolds number (based on the diameter) in the different domains for
the circular cylinder: (a) and (e) upstream in the side perspective, (b) and (f) upstream in
the top perspective, (c) and (g) downstream in the side perspective, and (d) and (h)
downstream in the top perspective. Solid line is the regression and dotted lines are the
95% confidence limits of the mean.

145

Figure 3.13: Flow visualization around Vallisneria americana using fluoroscein dye in
the side (left) and the top (right) perspectives. Velocities increase from top to bottom
panels [(a) and (b) are 1.3 cm s'\ (c) and (d) are 2.0 cm s'\ (e) and (f) are 5.0 cm s'\ (g)
and (h) are 8.4 cms ', and (i) and (j) are 11.0 cms ']. Flow is from left to right. Top and
side photos were taken sequentually.

146

Side

Top
Reynolds number (x 10^)

0.0

10

0.4

0.8

1.6

2.0

2.4

0.0

(a) y = (0.029+ 0.008)x + 0.33+0.05
F = 0.81, p = 0.034

8

0.4

(b)

0.8

1.2

1.6

2.0

2.4

y = (0.029+ 0.008)x + 0.310.8
F = 0.71, p = 0 .072

Side

6

I

1.2

Top

4
2

i
0

10

i 1

8

s

6

Io

4

P

_—--- " ------- -= j^ -

0

y = (0.22± 0.0 4)x + 3.0+0.3
= 0.89, p = 0.014

y = (0.1± 0.1)x + 4.5+0.9
F = 0.056, p = 0.69

(d)

h -

'
_----------------1

5

2
0

2 .0

(e )

y = (-0.05± 0.01)x + 0.63+ 0.09

y = (0.01± 0.03)x + 0.3+0.2
f = 0.077, p = 0.64

r = 0.79, p = 0.042

à 1.5
I& 1 .0
'o 0.5
§ 0 .0

i
a
8

I p

I

P

2 .0
(g )

1.5

(h)

y = (-0.07± 0.04)x + 0.8+0.2
= 0.54, p = 0.15

y = (-0,004± 0.015)x + 0.5±0.1
f = 0.024, p = 0.80

1 .0

0.5

t

__Î

0 .0

0

2

4

6

8

10

12

0

2

4

6

8

10

12

Velocity (cms'^)
Figure 3.14: Regression of area of dye coverage (a-d) and concentration of dye (e-h)
versus velocity or stem Reynolds number (based on the plant diameter) in the different
domains for Vallisneria americana: (a) and (e) upstream in the side perspective, (b) and
(f) upstream in the top perspective, (c) and (g) downstream in the side perspective, and
(d) and (h) downstream in the top perspective. Solid line is the regression and dotted lines
are the 95% confidence limits of the mean.

147

Figure 3.15: Flow visualization around Elodea canadensis using fluoroscein dye in the
side (left) and the top (right) perspectives. Velocities increase from top to bottom panels
[(a) and (b) are 1.3 cm s'\ (c) and (d) are 2.0 cms"\ (e) and (f) are 5.0 cms"*, (g) and (h)
are 8.4 cms *, and (i) and (j) are 11.0 cms *]. Flow is from left to right. Top and side
photos were taken sequentually.

148

Top

Side
Reynolds number (x 10 )
8

10

8

12

10

12

10

(a)

y = (-o.oi+o.oi)x + o.46 +o.o7

8

I

f = 0.29, p = 0.34

6
4

I

I

0

2
0

10

(c ) y = (-0.0004± 0.0004)x + 5.3±1.0

y = (0.10± 0.02)x + 3.810.2
= 0.84, p = 0.026

r^< 0.001, p = 0r99

8

I

( b ) y = (0.009+ 0.010)x + 0.63 1 0 .0 6
= 0.21, p = 0.43

1 6

I

4

Q

2
0

2 .0

(e)

y = (-0 .0 4 1 0 .0 2 )x + 0.6210.13
= 0.56, p = 0.14

(g )

y = (-0 .0 6 1 0 .0 2 )x + 0.7110.13

St/3
"o
§

0. 5

Ig

J- -

i §O
(/)

io

Q

» —» ft/io

(h )

y = (-0.0 6 1 0 .0 2 )x + 0.89 1 0 .1 2
= 0.78, p = 0.046

&
0 .5

0 .0
12
0_
2
4
6
8
10
12
Velocity (cms"')
Figure 3.16: Regression of area of dye coverage (a-d) and concentration of dye (e-h)
versus velocity or stem Reynolds number (based on the plant diameter) in the different
domains for Elodea canadensis: (a) and (e) upstream in the side perspective, (b) and (f)
upstream in the top perspective, (c) and (g) downstream in the side perspective, and (d)
and (h) downstream in the top perspective. Solid line is the regression and dotted lines are
the 95% confidence limits of the mean.
0

2

4

6

8

10

149

Figure 3.17: Flow visualization around Ceratophyllum demersum using fluoroscein dye
in the side (left) and the top (right) perspectives. Velocities increase from top to bottom
panels [(a) and (b) are 1.3 cms'*, (c) and (d) are 2.0 cm s'\ (e) and (f) are 5.0 cm s'\ (g)
and (h) are 8.4 cm s'\ and (i) and (j) are 11.0 cms"']. Flow is from left to right. Top and
side photos were taken sequentually.

150

Side
10

.0 4.5 9.0 13.5 18.0 22.5 27.0

(a)

0.0 4.5 9.0 13.5 18.0 22.5 27.0
(b )

y = (0.02± 0.01)x + 0.31±0.07
= 0.56, p = 0.14

8

Top

Reynolds number (x 10 )

y = (-0.01± 0.02)x + 0.5±0.1
= 0.14, p = 0.51

6

I

S'

4

I

I

I

<u

2
0
10
y = (-0.2± 0.1)x + 5.6±0.9
= 0.32, p = 0.31

y = (-0 .12± 0.03)x + 4.2±0.2
^ = 0.84, p = 0.028

fe l

y “ (-0 .001± 0.062)x + 0 .510.4

y = (-0 .0 7 1 0 .0 4 )x + 0 .9 1 0 .2
= 0.52, p = 0.16

(g )

y = (-0.07± 0.04)x + 0 .710.2
= 0.53, p = 0.16

8
6

K/l

4

I
Q

2
0
2 .0

0.5
0 .0

^

2 .0

g

Io

I-

(h)

y = (-0.0 3 1 0 .0 2 )x + 0.810.1
f = 0.48, p = 0.19

0.5
0 .0
0

2

4

6

8

10

12

8

10

12

Velocity (cms' )
Figure 3.18: Regression of area of dye coverage (a-d) and concentration of dye (e-h)
versus velocity or stem Reynolds number (based on the diameter) in the different
domains for Ceratophyllum demersum: (a) and (e) upstream in the side perspective, (b)
and (f) upstream in the top perspective, (c) and (g) downstream in the side perspective,
and (d) and (h) downstream in the top perspective. Solid line is the regression and dotted
lines are the 95% confidence limits of the mean.

151

CO

E
ü

i
Cylinder Vallisneria

Elodea Ceratophyllum

0 .8

co

E
ü
cn'
E

(b)
0 .6

0.4

CD

O)
C
CD 0 .2
JZ
O
O 0 .0

T

CD

c5 -0 .2

OC

Cylinder Vallisneria

Elodea Ceratophyllum

Cylinder Vallisneria

Elodea Ceratophyllum

0 .0 0

co

E
ü

-0 .0 2

3
O) -0.04
CD
C
CD
.C

o -0.06

0

1

oc -0.08

Figure 3.19: Comparison of the fluid dynamic conditions around plants and a physical
model (a) The velocity at which the transition from laminar to turbulent occurs for the
different plants; (b) Rate of change in the area of dye coverage with increasing velocity
for the different test subjects; and (c) The rate of change in the concentration of dye with
increasing velocity for the different test subjects. The dotted line represents the average
result for the physical model (circular cylinder).

152

Vertical velocity (w)

Streamwise velocity (u)
12

12

(h)
10

10

8
6

6

4

0

I

2

jKBu
g |||M

§

I
2^

I

Ù0

-4

0

—
0. 0
■■■■ 0.5
1.0

-2

2.0

_4

-0.4

f l

-0.2
0.0
0.2

-2

-2

12

12

(cl)

(c)
10

10

8

8

6

4

U

■ 8
■ 9
" 10
# 11
* 12
i£ 13

-2.0
-1.5

-1.0
-2

0.0

0.5

-4
-4

-4

-0.5

U
-2

2

4

Cross stream direction (y; cm)
Figure 3.20: Contours of the velocity field measured 3 cm from the flume bottom in a 2
cm X 2 cm grid in an empty flume. Each point of the contours was the mean of 120 s of
data samples at 25 Hz. (a) streamwise velocity (u) at 1.3 cms"\ (b) vertical velocity (w) at
1.3 cms'*, (c) streamwise velocity (u) at 11.0 cms'*, and (d) vertical velocity (w) at 11.0
cms Scale bars in cms'*.
Streamwise velocity (u)

Vertical velocity (w)

153

I n c r e m e n t # 2 - C y l in d e r (z)

12

12

(a)

(b)

10

10

8

8

6 -

4

2

0

i

§

-4
-4

a
M

1
2

J

u

-2

0 .0
0 .5
1 .0
1.5
2 .0

0
=

-0 .4

-2

mmwm 0.2
-4 -

12

(c)
10

h

00

8
6

4
2
0
-2

l:

-4

8

■

9

■ -1.5

10

■ - 1.0

11

" -0.5

2.0

12

0.0

13

0.5

-4

-2

-

-2

Cross stream direction (y; cm)
Figure 3.21: Contours of the velocity field measured 3 cm from the flume bottom in a 2
cm X 2 cm grid (3 cm in the vicinity of the cylinder) around a circular cylinder. Each
point of the contours was the mean of 120 s of data samples at 25 Hz. (a) streamwise
velocity (u) at 1.3 cms'*, (b) vertical velocity (w) at 1.3 cms'*, (c) streamwise velocity (u)
at 11.0 cms'*, and (d) vertical velocity (w) at 11.0 cms'*. Scale bars in cms'*.

Streamwise velocity (u)

Vertical velocity (w)

154

12

(a)
10

8
6

4
2

I

0

0.0

0.5
-2

U

§

1.5
2.0

-4

e

1.0

-4

-2

0(/)

12

1

(J)
10

(/3

I -0.5

Cross stream direction (y; cm)
Figure 3.22; Contours of the velocity field measured 3 cm from the flume bottom in a 2
cm X 2 cm grid (3 cm in the vicinity of the plant) around Vallisneria americana. Each
point of the contours was the mean of 120 s of data samples at 25 Hz. (a) streamwise
velocity (u) at 1.3 cm s'\ (b) vertical velocity (w) at 1.3 cm s'\ (c) streamwise velocity (u)
at 11.0 cm s'\ and (d) vertical velocity (w) at 11.0 cm s'\ Scale bars in cm s'\
Streamwise velocity (u)

Vertical velocity (w)

155

,2

(a)
,0 -

8
6 -

ib

:

4 .

2 ■
0

i
is
g

I

0.0
0 .5

-2

1.0

U

1 .5

2.0
-2

0.2

0

0c/5

12

1

(d)
10 -

on
8
6
4 -

2

—' 8
mmm 9

sa

0

— 10
" I

11

■

u

-2 f

12
mwMKi 13
-2

0

2

4

-

2.0

■ -1.5
■

- 1.0

I -0.5

I 0.0
i 0.5

-4

-2

Cross Stream direction (y; em)
Figure 3.23: Contours of the velocity field measured 3 cm from the flume bottom in a 2
cm X 2 em grid (3 cm in the vicinity of the plant) around Elodea canadensis. Each point
of the contours was the mean of 120 s of data samples at 25 Hz. (a) streamwise velocity
(u) at 1.3 cm s'\ (b) vertical velocity (w) at 1.3 cm s'\ (c) streamwise velocity (u) at 11.0
em s'\ and (d) vertical velocity (w) at 11.0 cms’'. Scale bars in cm s'\

Streamwise velocity (u)

Vertical velocity (w)

156

12

10
8

6

i

4

4

2

2

en

t

0

I
-2

IJ

g

0.0

%

t

0

0 .5

1.0

-2

-0 .4

0.2
0.0
0.2

U

1 .5

-

2.0

I

-4

2

12

-4

-2

-4

-4

12
(C )

( C l)

10

10

I
en

8
6

i

4

o

2

î

0
-2

IJ

8

11

-4

-2

2.0

-1 .5

10

-

-2

1.0

-0 .5

U

12
13

-4

-

9

0.0
0 .5

-4
-4

-2

2

4

Cross stream direction (y; cm)
Figure 3.24; Contours of the velocity field measured 3 cm from the flume bottom in a 2
cm X 2 cm grid (3 cm in the vicinity of the plant) around Ceratophyllum demersum. Each
point of the contours was the mean of 120 s of data samples at 25 Hz. (a) streamwise
velocity (u) at 1.3 cms"\ (b) vertical velocity (w) at 1.3 cm s'\ (c) streamwise velocity (u)
at 11.0 cm s'\ and (d) vertical velocity (w) at 11.0 cm s'\ Scale bars in cm s'\
Deflection
Oscillation

I

157

80

y =0X+0

y = 0 X+ 0

= -, p = -

60

0.8
0.6

40

0.4
20

0.2
0.0

0
-

0.2

80

(f)

60
40

§ 20

y = (0.05±0.02)x 4- 0.2 ±0.1

r = 0.67, p = 0.087

§

y = (0.3±0.2)x -h 56±1
= 0.25, p = 0.30

y

g
C/5

y

§

1

■■i
O
C/D
O

80

r = 0.96, p < 0.001

60

(p) y = (0.05±0.02)x -+- 0.1 ±0.1
= 0.68, p = 0.083

40

I

20

0

80

(h) y = (0.06±0.01)x + 0.04 ±0.09
r = 0.87, p = 0.018

60
40

20
y = (5.2±0.8)x -h 19 ±5

= 0.91, p = 0.003
0

2

4

6

8

10

12

0

8

10

12

Velocity (cms" )
Figure 3.25: Regressions for the angle of deflection for (a) the circular cylinder and the
different plant species (b) Vallisneria, (c) Elodea, and (d) Ceratophyllum. Regressions
for the number of oscillations per second with increasing velocity for (e) the circular
cylinder and the different plant species (f) Vallisneria, (g) Elodea, and (h)
Ceratophyllum. Solid line is the regression and dotted lines are the 95% confidence
limits of the mean.

158

7 1
(a)
CO

6 b
Ü
<D 5 03
c
cd 4 <D
03
c 3 cd
sz
Ü 2 o
■0)
+-> 1 cd
CC
0 Cylinder

Vallisneria

Elodea Ceratophyllum

Cylinder

Vallisneria

Elodea Ceratophyllum

0.08
CO
E 0.06
X

O) 0.04
CCS

ü

o 0.02

o
c5
X

0.00

Figure 3.26: (a) Rate of change in the angle of deflection with change in velocity for the
different test subjects; and (b) the rate of change in the number of oscillations per second
with change in velocity for the different test subjects. The cylinder is always at zero
because it was rigid and as such remained stationary.

159

a)

b)

c)

d)

e)

f)

Figure 3.27: The transition to turbulent flow of a fluid as it passes a circular cylinder seen
in cross section at different Reynolds numbers: (a) 1, (b) 10, (c) 13, (d) 26, (e)105, and (f)
150.
(From: http://nmm.media.mit.edu/student/95/aries/mas864/obstacles.html)

Chapter 4

160

The Effect of Internodal Spacing on Downstream Flow Patterns and Compliance in
Elodea canadensis and Ceratophyllum demersum
Abstract:
Aquatic macrophytes experience both diffusive and mechanical stresses, and as
such, must contend with different constraints. Specifically, at low velocities, plants must
overcome boundary layer limitations in order to increase exchange rates and hence
nutrient uptake, whereas at high velocities they must reduce drag to minimize tissue
damage or dislodgement. Given these constraints and the plasticity in macrophyte
morphology, the experimentally manipulated internodal spacing and compliance
(deflection of the plant from the vertical) of a whorled leaf macrophyte, Elodea
canadensis, and a whorled macrophyte with highly dissected leaves, Ceratophyllum
demersum, were examined in a flow chamber at velocities from - 1 - 1 1 cms '.
Fluoroscein dye was used to visualize the flow patterns and a digital camera was used to
record the images. Increasing the internodal spacing in Elodea affected the downstream
flow patterns by reducing the plant-flow interaction (e.g., less fluid came in contact with
the plant surface as compared to the other plants) but not the degree of compliance
exhibited by the plant. Conversely, in Ceratophyllum, the downstream flow patterns did
not change significantly with increased internodal spacing, but the compliance increased
considerably. These results suggest that Elodea may be better suited to faster flow
environments because it would experience less of a plant-flow interaction, which would
minimize hydrodynamic drag. Conversely, Ceratophyllum may be better suited to slower
flow environments because of the high plant-flow interaction, which would induce small
scale mixing close to the leaf surface, thus enhancing mass transfer. It appears that fluid

Chapter 4

161

dynamics have a determining role in the plastic response of macrophyte morphology, and
thus the evolution of plant form and function.
Introduction:
Freshwater macrophytes are phenotypically plastic in that they are able to change
their morphology and physiology in response to environmental conditions (IdestamAlmquist and Kautsky, 1995; Santamaria et al., 2003). This allows macrophytes to grow
in a wider range of environments than would be possible for a single morphotype
(DeWitt et al., 1998). Macrophytes are also able to change their morphology in response
to water motion and thus to colonize different hydrodynamic environments (Sculthorpe,
1967; Stewart and Carpenter, 2003). For example, at slower velocities entire leaf forms
in macroalgae tend to be shorter and thicker as opposed to longer and more slender in
faster velocities (Norton et al., 1982; Guitierrez and Fernandez, 1992) so that it would
seem logical to expect the same in freshwater macrophytes. Dissected morphologies
however, show the opposite trend whereby intemodal leaf segments are slender in slow
flow environments and shorter and firmer in faster moving water (Sculthorpe, 1967).
Macrophytes have also been known to increase their intemodal spacing in response to
low light levels (Cronin and Lodge, 2003), and to living in wave-sheltered sites (IdestamAlmquist and Kautsky, 1995). It would seem important, therefore, to determine whether
a change in plant architecture such as intemodal spacing effects the local fluid dynamic
conditions, and thus the diffusional and mechanical stress acting on submerged
freshwater plants.
Gas exchange and light capture are essential to the productivity of plants. One
way that plants enhance uptake, is through heterophylly (different leaf morphologies on

Chapter 4

162

the same plant, or in the same species, depending on environment), which is a prime
example of the functional explanation of plant architecture (e.g., Properpinaca
palusturis', Sculthorpe, 1967; Wells and Pigluicci, 2000). As a comparison, macroalgae
increase the surface area: volume in order to maximize light capture, and to increase the
flux of nutrients available to the plant (Hurd, 2000). Another way to increase nutrient
flux to the plant is through the possession of fine-scale roughness elements (such as fine
hairs, toothed edges, bullate surfaces etc.), which generate turbulent flow at the plant
surface at low velocities (Abelson et al., 1993; Hurd, 2000). One of the principle ways in
which this occurs is through vortex generation as the fluid passes by an object (Kundu,
1990; See chapter 3). This type of turbulence generation would also be important to
freshwater macrophytes as a means of increasing local mixing that would deliver
nutrients through inertial processes as opposed to diffusion (Okubo et al., 2002). It is
reasonable to postulate that intemodal spacing in freshwater macrophytes may be
important in the generation of vortices in that it would affect the presentation of whorls of
leaves to the mainstream flow.
Macrophytes must also be able to withstand the potential of mechanical damage
present in a flowing system due to fluid dynamic forces (e.g., lift, drag, shear stress,
acceleration forces; Vogel, 1994; Okubo et al., 2002). Flexibility can increase in the
ability of macrophytes to become compliant under flowing conditions, and thus decrease
the magnitude of hydrodynamic forces experienced (Sand-Jensen, 2003). Plants that
grow in energetic environments tend to be small in stature and limited in lateral spread,
whereas plants that grow in predictable habitats tend to be higher in stature and extensive
in lateral spread (Barrat-Segretain, 2001). In contrast, giant kelps are both long and

Chapter 4

163

flexible, and this morphology may reduce the drag experienced hy these algae (Denny et
al., 1997). It has also been found that macroalgae with fewer branches are better able to
withstand the mechanical stresses of higher velocities than highly branched macroalgae
(Sheath and Hamhrook, 1988). Schutten and Davy (2000) confirmed this result in
freshwater plants, where plants with less surface area were found to experience lower
drag. Clearly the ability of a plant to reduce the risk of damage is central to its survival
and, it would be useful to determine how freshwater macrophytes respond to increasing
flow conditions (i.e., remain stiff, comply, or break). As indicated above, intemodal
spacing is an important morphological feature that appears to confer benefits at both low
and high velocities. It would be instructive to investigate how this feature affects the
fluid dynamics surrounding freshwater macrophytes with whorled leaves. It is expected
that an increased intemodal spacing would generate less turbulence and less compliance
by the plant. The following chapter examines the null hypothesis that intemodal spacing
has no effect on the flow or compliance of whorled-leafed macrophytes with entire and
highly dissected leaves (i.e., Elodea canadensis and Ceratophyllum demersum,
respectively).
Materials and Methods:
In order to examine the null hypotheses, photographs and short videos were taken
of Fluoroscein dye moving: (1) in an empty test section; (2) past a circular cylinder (0.7
cm diameter); (3) past Elodea canadensis (referred to as Elodea in this report); (4) past
Elodea with the leaf whorl at every second node removed; (5) past Elodea with the leaf
whorl at every second and third node removed; (6 ) past Ceratophyllum demersum
(referred to as Ceratophyllum in this report); (7) past Ceratophyllum with the leaf whorl

Chapter 4

164

at every second node removed; and (8 ) past Ceratophyllum with the leaf whorl at every
second and third node removed (all plant segments in 5 cm lengths: Figures 4.1, 4.3, 4.5,
4.7,4.9,4.12,4.15). All specimens of Elodea were collected from Tabor Lake, Prince
George, B.C., and all specimens of Ceratophyllum were purchased from Ward Scientific
Supply (Rochester, NY, USA). The whorls were removed where they attached to the
stem by pulling them off with tweezers. A sample size of three was used for each intact
and manipulated segment of each plant.
The experimental protocols and procedures were the same as those used in
Chapter 3. The plants were examined in a recirculating flow chamber (see Chapter 3;
Figure 3.1) operated at velocities ranging from about 1-11 cms"\ similar to the velocities
used by Nepf and Koch (1999), and Leonard and Luther (1995). Fluoroscein dye was
injected isokinetically upstream of the plant and excited by a black light. A white light
was also used to ensure that the plants were visible and could be recorded as photographs
and video clips by the digital camera (Nikon CoolPix995; Melville, NY, USA). A pixel
intensity versus dye concentration curve was used in the image analysis undertaken in
MATLAB version 6.5 (Mathworks, Natick, Massachusetts). The area of dye coverage
and the concentration of dye were determined for an upstream and downstream domain
with respect to the plant in both the side and top perspective. The angle of deflection
(angle from the vertical), the frequency of oscillations, and the angle of the leaves of
Ceratophyllum from the stem were also measured from the video recordings.

Chapter 4

165

Statistical Analysis
Effect of Velocity on Dye Coverage and Concentration
The average and standard deviation of the area of dye coverage and/or the
concentration of dye in that area, were computed from the three images taken at each
velocity. The area or concentration was then regressed on the test velocity and whorl
Reynolds number based on the diameter of the subject (i.e.,

_ul
^ ’ ^here 1 is the

diameter of the stem plus the leaves, u is the velocity, and t) is the kinematic velocity;
Vogel, 1994). The null hypothesis examined was that the slope of the regression is equal
to zero, in other words there is no trend in area of dye coverage or concentration of dye
with increasing velocity.
Effect of Intemodal Spacing on the Rate of Oscillation
Analyses of covariance (ANCOVAs) were conducted using the frequency of
oscillations as the dependent variable. The different plants were the categorical factors,
and velocity was the continuous predictor (the covariate). A Fisher LSD test was used
for Post Hoc comparisons. This statistical test was chosen to test whether there is a
difference between slopes for the different regressions.
Effect of Velocitv and Node Position on the Leaf Angles of Ceratophyllum
The dissected nature of the leaves of Ceratophyllum provided the opportunity to
examine the response of leaf segments on the upstream and downstream side of the plant.
The leaves were defined as moving closer to the stem when they moved towards the stem
in the same direction as the flow in the chamber. ANCOVAs were conducted using angle
of the Ceratophyllum leaf segment from the stem as the dependent variable. The node
position on the stem was the categorical factor, and velocity was the continuous predictor

Chapter 4

166

(the covariate). A Fisher LSD test was used for Post Hoc comparisons between leaf
angles of the different nodes.
Results:
It was established in Chapter 3 that the upstream flow patterns were not
significantly different in any perspective or morphology because the dye was injected
isokinetically. Consequently, only the downstream flow patterns are examined here.
Cylinder
The flow patterns downstream of the circular cylinder have been discussed in the
previous chapter, but are included here for comparative purposes. Laminar flow was
observed at the slower velocities, as indicated by the loops of dye (“stickiness”), and the
large area of recirculation downstream of the cylinder. The transition from laminar to
turbulent flow occurred between 5.0 cms'^ and 8.4 cms'^ (i.e.. Re ~ 350-590). At the
faster velocities, the flow was turbulent, with fewer loops of dye and a smaller
recirculation zone. This is best seen in the top perspective where the dye streaklines
revealed a classic attached eddy at 1.3 cms'^ (Re ~ 90; Figure 4.1b), a von Karman vortex
street at 2 cms ' (Re ~ 140; Figure 4. Id), and more turbulent structures (eddies) at higher
velocities (Figure 4.1f,h,J).
There was a significant increase in the dye coverage downstream in the side
perspective with velocity (i.e., 0.5 cm^/cms ^ p = 0.018; Figure 4.2c), and a positive trend
in the top perspective (p = 0.41; Figure 4.2d). Additionally, the dye concentration tended
to decrease with velocity in the up and downstream domains, and this was significant in
the case of the downstream side perspective (p = 0.014; Figure 4.2g).

Chapter 4

167

Elodea
The pattern of flow around Elodea was reported in Chapter 3 but it is repeated
here for comparative purposes. The downstream dye streaklines were attached to the
individual whorls in filamentous loops, and recirculation was present at the lower
velocities (Figure 4.3a-d). The transition from laminar to turbulent occurred at ~ 5.0 cms'
^ (Re ~ 500), as indicated by the onset of eddies downstream of the plant (Figure 4.3e,f ).
Von Karman vortices were shed at 8.4 cms'^ (Re ~ 840; Figure 4.3h) and more turbulence
(widely spaced eddies, and less stickiness and recirculation) was evident at 1 1 . 0 cms"'
(Re ~ 1100; Figure 4.3i-j).
There was a significant increase in the area of dye coverage with velocity
downstream in the top perspective (p = 0.026) of ~ 0.10 cm^/cms"' (Figure 4.4d).
Additionally, the dye concentration tended to decrease with velocity in the up and
downstream domains, and this was significant in the case of the downstream side (p =
0.049) and top perspectives (p = 0.046; Figure 4.4g,h).
Elodea with the Leaf Whorl at Everv Second Node Removed
The dye streaklines downstream of Elodea with the whorls at every second node
removed showed a large amount of “stickiness” to the whorls at the lower velocities
(Figure 4.5a-f). The transition from laminar to turbulent occurred between 5.0 cms"' and
8.4 cms"' (Re ~ 500-840; Figure 4.5e-h). At the higher velocities, turbulence was
indicated by the closely spaced eddies, and the reduced amount of stickiness and
recirculation as compared to the slower velocities (4.5g-j). There was also some
compliance (bending into the flow) with increasing velocity.
There was a significant increase in the area of dye coverage with velocity
downstream in the side perspective (p = 0.015), which was on the order of ~ 0.2

Chapter 4

168

cm^/cms'^ (Figure 4.6c), although this trend was not significant in the top perspective.
Additionally, the dye concentration tended to decrease with velocity in the upstream and
downstream domains, and this was significant in the case of the downstream top
perspective (p = 0.016; Figure 4.6g,h).
Elodea with the Leaf Whorls at Everv Second and Third Node Removed
The dye streaklines downstream of Elodea with the whorls at every second and
third node removed were laminar, as indicated by the loops in the dye attached to the
individual whorls, at the lower velocities (Figure 4.7a,b). The transition from laminar to
turbulent flow occurred between 5.0 ems'* and 8.4 cms'* (Re ~ 500-840) as indicated by
the dye streakline becoming increasingly wavy (Figure 4.7c-h). The onset of eddies
denotes that the flow was turbulent at the higher velocities (Figure 4.7g-j). There was
little compliance, or bending into the flow with increasing velocity.
There were no significant trends in the area of dye coverage with velocity
downstream of the plant, and the trend in the side perspective was positive, but was
negative in the top perspective (Figure 4.8c,d). The dye concentration however,
decreased significantly with velocity in the downstream side (p = 0.035) and top (p =
0.007) perspectives at -0.07 and-0.05 gL'*/cms'*, respectively (Figure 4.8g,h).
Ceratophyllum
The flow patterns downstream of Ceratophyllum were discussed in Chapter 3, and
are presented here for comparative purposes. The dye streaklines downstream of
Ceratophyllum “stuck” to the individual leaf segments at 1.3 cms'* (Re ~ 300), and there
was considerable recirculation behind the plant (Figure 4.9a,b). The transition of the dye
from smooth laminar to eddies occurred between 5.0 and 8.4 cms * (Re ~ 1200-2000;

Chapter 4

169

Figure 4.9c-f). Closely spaced eddies, and the reduced amount of stickiness and
recirculation, indicated turbulent flow at the higher velocities (Figure 4.9g-j).
There was a significant decrease in the area of dye coverage with velocity
downstream in the top perspective (p = 0.028), with a rate on the order of -0 . 1 cm^/cms"'
(Figure 4.10d). There was also a tendency for the dye concentration to decrease with
velocity in the up and downstream domains, although no cases were significant (Figure
4.10e-h).
An examination of the effect of velocity on leaf angle on the upstream side of the
stem revealed that all leaves moved closer to the stem with increasing velocity (~ l°/cms"*), although this was only significant for the bottom most (p = 0.006) and the third
leaf up from the bottom (p = 0.02; Figure 4.lla,c). On the downstream side of the stem,
the bottom three leaves moved significantly closer to the stem with increasing velocity (~
-l°/cms'^), but moving up the stem the trend was similar for the next two leaves, but not
significant, and the leaf at the apex moved away from the stem, though the trend was not
significant (Figure 4.11a-f).
Ceratophyllum with the Leaf Whorls at Everv Second Node Removed
The dye streaklines downstream of Ceratophyllum with the whorls at every
second node removed were laminar, as denoted by smooth dye streaklines and the large
amount of “stickiness” on the individual leaf segments at the lower velocities. At the
lower velocities, there was also recirculation behind the plant (Figure 4.12a-f). The
transition from laminar to turbulent flow occurred between 5.0 cms'^ and 8.4 cms'^ (Re ~
1 2 0 0 -2 0 0 0 ) as indicated by the dye streakline changing from smooth to more chaotic in

structure (Figure 4.12e-h). At the higher velocities, turbulence was indicated by the

Chapter 4

170

closely spaced eddies, and less stickiness and recireulation (Figure 4.12g-j). There was
also a high amount of compliance of the plant into the flow with increasing velocity.
There was a positive but not significant trend in the area of dye coverage with
velocity downstream of the plant in the top perspective (Figure 4.13d). There was
however, a significant decrease in dye concentration of -0.07 gUVcms'^ (p = 0.039) in
the side and -0.031 gUVcms'* (p = 0.007) in the top perspective (Figure 4.13g,h).
The leaf at the base of the plant on the upstream side tended to move away from
the stem with increasing velocity, but this was not significant. Conversely, the next leaf
showed a non-significant trend to move closer to the stem, and the top three leaves
moved significantly closer to the stem with increasing velocity (i.e., -1.6 to -3.2 °/cm s'\
Figure 4.14a-e). On the downstream side of the stem, the leaf at the base of the plant
tended to move towards the stem, the next leaf moved significantly away from the stem at
a rate of ~ 0.5 7cms '. The third and fourth leaves from the bottom of the plant moved
towards the stem with increasing velocity, but the leaf at the apex of the plant moved
significantly away from the stem with increasing velocity (Figure 4.14a-e).
Ceratovhvllum with the Leaf Whorls at Everv Second and Third Node Removed
At the slower velocities, laminar flow was indicated by the filamentous dye loops
attached to the individual leaf segments, and the recirculation behind Ceratophyllum with
the whorls at every second and third node removed (Figure 4.15a-f). The dye streakline
downstream became increasingly complex with increasing velocity and the transition
occurred between 5.0 cms"^ and 8.4 cms'^ (Re ~ 1200-2000; Figure 4.15e-h). At the
higher velocities, turbulence was indicated by the closely spaced eddies, and a reduced

Chapter 4

171

amount of stickiness and recirculation (Figure 4.15g-j). This manipulated segment of
Ceratophyllum showed a high amount of compliance with velocity, similar to above.
There was no significant trend in the area of dye coverage downstream in the side
perspective although there was a tendency for an increase with velocity (Figure 4.16c,d).
There was also a tendency for the dye concentration to decrease with velocity in both the
side and top perspectives (Figure 4.16g,h).
The leaf at the base of the plant tended to move closer to the stem as did the
second and third leaves up the stem, and the leaf at the apex did so significantly (p =
0.031, 0.026, <0.001, respectively). The rate of change in compliance towards the stem
increased as the leaves approached the apex of the plant (-1.3 to -2.0 7cms“^; Figure
4.17a-d). On the downstream side of the stem, the bottom three leaves tended to move
away from the stem, whereas the leaf at the apex tended to move closer to the stem with
increasing velocity (Figure 4.17a-d).
Effect of Manipulating Intemodal Spacing
Area o f Dye Coverage
An ANCOVA was undertaken on all of the plant types to determine whether there
were significant differences among their responses to velocity. There was a significant
effect of velocity on area of dye coverage (p = 0.023; Table 4.1a), plant type (i.e., species
and intemodal spaces; p < 0.001), side versus top perspective (p = 0.036), and upstream
versus downstream domain (p < 0.001). There were also significant interactions
between the plant and the upstream versus downstream domain (p = 0 .0 0 1 ), as well as
between the side versus top perspective and the upstream versus downstream domain (p =
0 .002).

Chapter 4

172

Upon closer inspection however, ANCOVAs conducted on each plant species and
manipulation revealed that in terms of area of dye coverage, there was a significant effect
of velocity for the cylinder (p < 0.001; Table 4.2a), but not for Elodea (p = 0.52; Table
4.2b), or Elodea with the whorl at every second node removed (p = 0.08; Table 4.2c), or
for Elodea with the whorls at every second and third node removed (p = 0.78; Table
4.2d). There were significant differences between the upstream and downstream domains
for all plants, but no significant effect of side versus top perspectives for any of the
manipulations (Table 4.2). Pairwise comparisons in the side perspective downstream
revealed significant differences in the area of dye coverage between the cylinder and
Elodea with the whorls at every second and third node removed (p = 0.001), the cylinder
and the empty test section (p = 0.003), Elodea and Elodea with the whorls at every
second and third node removed (p < 0.001), Elodea and the empty test section (p =
0.001), and between Elodea with the whorls at every second node removed and Elodea
with the whorls at every second and third node removed (p = 0.039: Table 4.3c). In the
top perspective, significant differences occurred between the subjects and the empty test
section (p < 0.001), but not among any of the subjects (Table 4.3d).
ANCOVAs were conducted on Ceratophyllum and its manipulated stems. This
analysis demonstrated that there was only a marginal effect of velocity on the area of dye
coverage for Ceratophyllum (p = 0.08; Table 4.4b), and no effect for Ceratophyllum with
the whorl at every second node (p = 0.39; Table 4.4c) and every second and third node
removed (p = 0.11; Table 4.4d). There was however, a significant upstream versus
downstream effect, but no significant effect of side versus top perspective for either of
the plants (Table 4.4). Pairwise comparisons showed that the area of dye coverage in the

Chapter 4

173

side perspective was significantly different between Ceratophyllum with the whorls at
every second node removed and the cylinder (p = 0.006), the cylinder and the empty test
section (p = 0.003), and between Ceratophyllum and the empty test section (p = 0.042:
Table 4.5c). In the top perspective, significant differences in area of dye coverage
occurred between the subjects and the empty test section, the cylinder and Ceratophyllum
(p = 0.023), and between the cylinder and Ceratophyllum with the whorls at every second
node removed (p = 0.007; 4.5d).
Concentration o f Dye
An ANCOVA was conducted on the concentration of dye for all of the subjects,
which revealed a significant effect of velocity (p < 0.001; Table 4.1b), plant type (p =
0.39), side versus top perspective (p = 0.007), and upstream versus downstream domain
(p < 0.001). Separate ANCOVAs conducted for Elodea and its manipulated intemodes
showed that there was a significant effect of velocity on the concentration of dye for all
of the plants (Table 4.2). However, there were no significant effects of upstream versus
downstream domain, or side versus top perspective for any of the plants (Table 4.2).
Pairwise comparisons showed that in the side perspective, downstream of the Elodea
manipulations, significant differences occurred between the subjects and the empty test
section (Table 4.6c), but not among any of the subjects. In the top perspective, the only
significant difference occurred between the cylinder and the empty test section (p =
0.029; Table 4.6d).
ANCOVAs conducted on the concentration of dye for the different
Ceratophyllum intemodal manipulations showed that velocity had a significant effect on
Ceratophyllum with the whorl from every second node removed (p = 0.003; Table 4.4c)

Chapter 4

174

and Ceratophyllum with the whorls from every second and third node removed (p <
0.001; Table 4.2d), but only marginal significance for unmanipulated section of
Ceratophyllum (p = 0.059; Table 4.4b). The same pattern occurred in the upstream
versus downstream position, but no significant effects were noted between the side and
top perspectives (Table 4.4). In the side perspective, downstream of the Ceratophyllum
manipulations, significant differences in concentration were found between the cylinder
and Ceratophyllum (p = 0.040), the cylinder and the empty test section (p = 0.002), and
between Ceratophyllum and the empty test section (p = 0.020: Table 4.7c). In the top
perspective downstream, significant differences were found between the cylinder and
Ceratophyllum (p = 0.031), Ceratophyllum with the whorls at every second and third
node removed (p = 0.004), and the empty test section (p = 0.029; Table 4.7d).
Angle o f Leaf from the Stem
An ANCOVA of the leaf angle results for the different manipulations of
Ceratophyllum revealed a significant effect of manipulation on leaf angle (p = 0.005:
Table 4.8a). However, there were only significant differences between the leaves on the
upstream side of the stem of Ceratophyllum, and the leaves in the upstream and
downstream position on the manipulated segments of Ceratophyllum (Table 4.8b). In
terms of the rate of change in the leaf angle, the apical leaf in the approaching flow
moved towards the stem at a faster rate than the leaves lower down on the stem for all
three plant segments (Figure 4.18a). Only unmanipulated Ceratophyllum showed a trend
in the downstream direction, whereby the apical leaf moved away from the stem, and the
other leaves moved towards the stem at a faster rate the closer they were to the base of
the plant (Figure 4.18b). Neither of the Ceratophyllum manipulations showed a trend in

Chapter 4

175

the rate of change in leaf and angle from the stem with increasing velocity. The same
analyses were not conducted for Elodea due to the relatively short and stiff structure of
the leaves, which were not observed to change in angle of deflection upon viewing the
video clips.
Angle o f Deflection
The angle of deflection for the different species and their associated
manipulations increased significantly with increasing velocity (p < 0.003; Table 4.19a-g).
The rate ranged from 2-3 °/cms'^ for Elodea and 4-5 °/cms'^ for Ceratophyllum. As such,
the ANCOVA showed that there was a significant effect of velocity (p < 0.001), as well
as for plant type (p < 0.001; Table 4.9a). More specifically, post hoc tests showed that
there were significant differences between the cylinder and Elodea, and the three
Ceratophyllum manipulations. Differences also occurred: between Elodea and Elodea
with the whorls at every second node removed and Ceratophyllum', between Elodea with
the whorls at every second and third node removed and the three Ceratophyllum
manipulations; Elodea with the whorls at every second and third node removed and
Ceratophyllum and Ceratophyllum with the whorls at every second and third node
removed; and between unmanipulated Ceratophyllum and the two Ceratophyllum
manipulations (Table 4.9b).
Oscillations
The frequency of oscillations increased with velocity for both plant species and
their associated manipulations (Figure 4.20a-g). This was significant for Elodea with the
whorls at every second node removed (p = 0.031), Ceratophyllum (p = 0.018),
Ceratophyllum with the leaves at every second node removed (p = 0.006), and

Chapter 4

176

Ceratophyllum with the leaves at every second and third node removed (p = 0.048). The
ANCOVA showed that velocity (p < 0.001) and plant type (p = 0.002; Table 4.10a) had a
significant effect on the frequency of oscillation. However, pairwise comparisons
revealed that all of the significant pairwise differences occurred between the cylinder
(which did not oscillate) and the test subjects (Table 4.10b).
Comparison o f Plant Types
The transition from laminar to turbulent flow, occurred at a velocity equal to or
greater than the physical model for all plants with the exception of unmanipulated Elodea
(Figure 4.21a). The transition for the cylinder, Elodea with the leaves from every second
node removed, and Ceratophyllum occurred between 5 and 8.4 cm s'\ for Elodea between
2 and 5 cm s'\ at ~ 8.4 cms'^ for Elodea with the leaves from every second and third node
removed, and between 8.4 and 11 cms”' for Ceratophyllum with the leaves from every
second, and every second and third node removed. It should be noted that the use of Re
did not change the pattern of observation reported here. Specifically the Re for the
cylinder ranged from 90-770, from 130-1100 for Elodea and its manipulations, and from
300-2600 for Ceratophyllum and its manipulations. Collectively, this would indicate that
the manipulations reduced the level of local mixing.
A direct comparison of the rate of change in the area of dye coverage with
increasing velocity revealed that the rate was slower for all of the plants than for the
physical model (Figure 4.21b). The rate of change for the circular cylinder increased
significantly on the order of ~ 0.5 cm^/cms"\ as it did for Elodea with the whorl at every
second node removed although at a slower rate on the order of ~ 0.2 cm^/cms‘\ There
was a positive, although non-significant trend seen in Elodea with the whorls at every

Chapter 4

177

second and third node removed (~ 0.08 cm^/cms'^), and in Ceratophyllum with the whorls
at every second and third node removed (~ 0.2 cm^/cms'^). Non-significant negative
trends in the area of dye coverage with increasing velocity were seen in Elodea (~ 0.0004 cm^/cms ’), Ceratophyllum (~ -0.2 cmVcms"’), and Ceratophyllum with the whorl
at every second node removed (~ -0 . 0 1 cm^/cms'^).
A direct comparison o f the rate of change in concentration o f dye with velocity
showed that the rate of change for the plants was less than the rate for the physical model
(Table 4.21c). The rate for the circular cylinder decreased significantly, at ~ -0.05 gL
V cm s'\ as did Elodea (~ -0.06 gL'Vcms'^), Elodea with the whorls at every second and
third node removed (~ -0.07 gL'Vcms'*), and Ceratophyllum with the whorl at every
second node removed (~ -0.07 gUVcms'^). The rates for Elodea with the whorl at every
second node removed (~ -0.06 gL'Vcms'^), Ceratophyllum (~ -0.07 gL'Vcms'^), and

Ceratophyllum with the whorls at every second and third node removed (~ -0.08 gL
Vcms'^), also decreased though non-significantly.

The rate of change in the angle of deflection with increasing velocity was lower
for the Elodea series than for the Ceratophyllum series of manipulations (Figure 4.22a).
Elodea and Elodea with the whorl at every second node removed showed an increase on
the order of - 3 . 1 °/cm s'\ whereas Elodea with the whorls at every second and third
node removed showed a much slower rate of change at - 1.9 °/cm s'\ The rate of change
for Ceratophyllum (~ 5.2 7cms'^), Ceratophyllum with the whorl at every second node
removed (~ 4.4 7cms'^), and Ceratophyllum with the whorls at every second and third
node removed (- 5.2 7cms'*) were similar. The rate of change in the frequency of

Chapter 4

178

oscillations with increasing velocity was similar for all plants although the Elodea
manipulations oscillated at lower rates (i.e., 0.05 vs. 0.07 Hz/cms'^; Figure 4.22b).
Discussion:
The response of flow to intemodal spacing was different between the entirewhorled (Elodea) and dissected-whorled (Ceratophyllum) morphologies. For Elodea the
plant-flow interaction decreased with increased spacing, reducing the opportunity for dye
attachment (Schwinge et al., 2002), and thus exchange between the fluid and the leaf
surface. For Ceratophyllum and the associated manipulations, the plant-flow interaction
was greater because the angle of the leaf from the stem did not differ significantly among
the different manipulations. Resultantly, there was always a highly dissected, mesh-like
obstruction maintaining similar downstream flow patterns (i.e., small scale eddies)
(Nowell and Jumars, 1984; Sand-Jensen and Pedersen, 1999). This would enhance the
local mixing and contact of the fluid with the leaf surface, a feature that suggests this
morphology may be better suited to slower flows.
Despite differences in dye patterns, the concentration of dye decreased at a slower
rate for the plants than for the cylinder. This suggests that although plants, and different
manipulations thereof, have different ways of affecting the flow, they still retained more
fluid than a physical model. This would provide more opportunity for nutrient uptake. A
similar trend was demonstrated in macroalgae by Hurd et al. (1996) where different blade
morphologies produced different flow patterns, but the uptake of nitrogen remained
similar between the two morphologies. It is not known whether other nutrients would
behave the same manner. It seems reasonable to expect that a variation in the
morphology of a freshwater plant species (i.e., intemodal spacing) and the resultant fluid

Chapter 4

179

dynamic patterns identified above may influence nutrient uptake. This is an issue that
should to be examined experimentally in freshwater macrophytes.
The compliance of Elodea and Ceratophyllum increased with velocity which
would increase the interception efficiency of both direct and diffuse light but could lead
to canopy shading (Niinemets and Fleck, 2002). In addition, there was a trend for the
individual leaves of the Ceratophyllum to fold in closer to the stem as velocity increased,
which would decrease drag, but also photosynthesis due to self-shading. Subsequently, at
slower velocities where diffusion is limiting, the leaves protrude more into the flow, thus
presenting more surface area for photon capture and increasing the efficiency of light use
by the plant (Gutschick, 1999; Niinemets and Fleck, 2002). Leaf shape, the length of
intemodes, and the pattern of leaf arrangement are all important in defining the total
amount of direct light a shoot can receive (Niklas, 1992). This is especially true for
aquatic macrophytes because water attenuates the intensity of sunlight and preferentially
absorbs the red wavelengths of light, so that submerged aquatic plants experience
significantly lower light intensities shifted in favour of blue wavelengths (Niklas, 1997).
These results demonstrate that an increasing intemodal spacing has different
implications for different plant morphologies. For example, Elodea may be better suited
to faster flow because the plant-flow interaction decreased with increased intemodal
spacing. Conversely, Ceratophyllum would probably be better suited to slower flow
conditions because the highly dissected morphology ensures greater plant-flow
interaction. Additionally, the high level of compliance would increase the amount of
surface area exposed for photon capture as well as reduce drag, as mentioned above.

Chapter 4

180

It is important to recognize that the plants in this study were grown in a low flow
environment prior to the experiments. Had they been taken from high flow areas, there
may have been morphological differences such as shorter intemodal spaces (IdestamAlmquist and Kautsky, 1995) and more robustness in their stmcture (Schutten and
Davey, 2000). This however, should have minimal effects on the results of the
experiment because the plant segments were experimentally manipulated to alter the
intemodal spacing, and since the experiments were conducted at the lower end of the
velocity range, breakage was not an issue, although there may have be a decrease in the
degree of compliance.
Often, while plants acclimate to increase productivity, they increase the chance of
mechanical damage (Ennos, 1999). For example, a natural response of aquatic
macrophytes to increase photosynthesis is to grow closer to the light source (i.e., away
from the bottom), thus increasing mechanical stress by exposing the plant to higher
velocities (Ennos, 1999; Strand and Weisner, 2001). A strategy of minimizing
rnechanical stress in aquatic systems is for plants to become more compliant with the
flow (streamlined) in order to reduce the hydrodynamic drag (Schulthorpe, 1967).
However, in order to inhabit faster moving environments, the length of the leaf must be
small since drag is a function of surface area (Vogel, 1994). Hence, in order to reduce
drag, plants must either become compliant to the flow, or reduce the size of individual
surfaces facing the flow. Therefore, an increase in intemodal spacing would be
beneficial to plants that grow higher in the water column in relatively low flow
environments beeause the reduction in leaf area would decrease drag. This is important
because the plants need to remain erect in the water column in order to maximize light

Chapter 4

181

interception. An increase in intemodal spacing would not be beneficial at higher
velocities because the plants would either be deflected downwards (i.e., increased
compliance) and/or not be robust enough to withstand high drag forces.
Intemodal spacing is a form of phenotypic plasticity that can link certain plant
characteristics to environmental conditions (Santamaria et al., 2003; Idestam-Almquist
and Kautsky, 1995). For example, plants growing in wave-protected sites tend to be
longer, with longer intemodes as well as more shoots with longer branches, as compared
to those in the wave-exposed sites (Idestam-Almquist and Kautsky, 1995). This pattem
also applies to marine invertebrates where intersetal spacing increases with decreasing
velocity (Sebens et al., 1997; Marchinko and Palmer, 2003). It would appear that this
form of phenotypic plasticity (i.e., intemodal spacing and compliance) is important for
aquatic macrophytes.
The null hypothesis that intemodal spacing has no effect on the flow patterns
generated by the plants and their compliance under different velocities was rejected.
Intemodal spacing has been found to affect surface area to volume ratios (Denny, 1993;
Hurd and Stevens, 1997; Gutschick, 1999) and the degree of compliance (Sculthorpe,
1967; Sand-Jensen, 2003; Speck, 2003). The surface area:volume ratio is important for
diffusion, local mixing, and drag. For example, the aerial form of the heterophyllous
plant, Properpinaca palustris, has leaves with toothed margins that are smaller, broader,
thicker and fewer in number than those on submerged streams, which are highly dissected
(Schulthorpe, 1967; Wells and Pigluicci, 2000). This high level of dissection is thought
to increase the surface arearvolume ratio, but the small leaf segments would also work to
trip boundary layers (i.e., change them from laminar to turbulent) (Gutschick, 1999), thus

Chapter 4

182

increasing the delivery of gases to the leaf surface. In addition to plants altering their
morphology in response to environmental conditions, plants can also change their local
conditions, as was shown by Koehl and Alberte (1988), who found that the flapping of
macroalgae increased photosynthesis, whereas the clumping together of fronds decreased
photosynthesis due to self shading. This demonstrates that the macrophyte-flow
interaction is a complicated one in that water motion has an effect on plant morphology,
and this interaction can affect productivity. In other words, fluid dynamics have had a
determining role in the plasticity of freshwater macrophytes because of the plant-flow
interaction, which matches plant morphology to environmental condition. In this case,
increased intemodal spacing allowed the plant to remain higher in the water column in a
relatively low energy environment because of a lower plant-flow interaction.

Chapter 4

183

Literature Cited
Abelson, A., T. Milch, and Y. Loya. 1993. Flow patterns induced by substrata and body
morphologies of benthic organisms, and their roles in determining availability of food
particles. Limnology and Oceanography. 38(6): 1116-1124.
Barrat-Segretain, M-H. 2001. Biomass allocation in three macrophyte species in relation
to the disturbance level of their habitat. Freshwater Biology. 46:935-945.
Cronin, G., and D.M. Lodge. 2003. Effects of light and nutrient availability on the
growth, allocation, carbon/nitrogen balance, phenolic chemistry, and resistance to
herbivory of two freshwater macrophytes. Oecologia. 137:32-41.
Denny, M.W. 1993. Air and Water The Biology and Physics o f Life’s Media. Princeton
Univeristy Press, U.S.A. 341 pp.
Denny, M.W., B.P Gaylord, and E.A. Cowen. 1997. Flow and Flexibility II. The roles of
size and shape in determining wave forces on the bull kelp Nereocystis luetkeana. The
Journal o f Experimental Biology. 200:3165-3183.
DeWitt, T.J., A. Sih, and D.S. Wilson. 1998. Costs and limits of phenotypic plasticity.
TREE 13(2):77-81.
Ennos, A.R. 1999. The aerodynamics and hydrodynamics of plants. The Journal o f
Experimental Biology. 202:3281-3284.
Gutierrez, L.M., and C. Fernandez. 1992. Water motion and morphology in Chondrus
crispus (Rhodothyta). Journal o f Phycology. 28:156-162.
Gutschick, V.P. 1999. Research reviews: Biotic and abiotic consequences of differences
in leaf structure. New Phytologist. 143(1):3-18.
Hurd, C.L. 2000. Water motion, marine macroalgal physiology, and production. Journal
o f Phycology. 36:453-472.
Hurd, C.L., P.J. Harrison, and L.D. Druehl. 1996. Effect of seawater velocity in
inorganic nitrogen uptake by morphologcially distinct forms of macrocystis intgrifolia
from wave-sheltered and exposed sites. Marine Biology. 126:205-214.
Hurd, C.L. and C.L. Stevens. 1997. Flow visualization around single- and multiplebladed seaweeds with various morphologies. Journal o f Phycology. 33:360-367.
Idestam-Almquist, J., and L. Kautsky. 1995. Plastic responses in morphology of
Potamogeton pectinatus L. to sediment and above-sediment conditions at two sites in
the northern Baltic proper. Aquatic Botany. 52:205-216.

Chapter 4

184

Koehl, M.A.R., and R.S. Alberte. 1988. Flow, flapping, and photosynthesis of
Nereocystis luetkeana: a functional comparison of undulate and flat blade
morphologies. Marine Biology. 99:435-444.
Kundu, P.K. 1990. Fluid Mechanics. Academic Press, San Diego, Toronto. 638 p.
Marchinko, K.B., and A.R. Palmer. 2003. Feeding in flow extremes: dependence of cirrus
form on wave-exposure in four barnacle species. Zoology. 106:127-141.
Nepf, H.M., and E.W. Koch. 1999. Vertical secondary flows in submersed plant-like
arrays. Limnology and Oceanography. 44(4): 1072-1080.
Niinemets, U., and S. Fleck. 2002. Petiole mechanics, leaf inclination, morphology, and
investment in support in relation to light availablility in the canopy of Liriodendron
tulipifera. Oecologia. 132:21-33.
Niklas, K.J. 1992. Plant Biomechanics An Engineering Approach to Plant Form and
Function. The University of Chicago Press, Chicago, 607 pp.
Niklas, K.J. 1997. The Evolutionary Biology o f Plants. The University of Chicago Press,
Chicago, 449 pp.
Norton, T.A., A C. Mathieson, and M. Neushul. 1982. A review of some aspects of form
and function in seaweeds. Botanica Marina. XXV: 501-510.
Nowell, A R M., and P.A. Jumars. 1984. Flow environments of aquatic benthos. Annual
Review o f Ecology and Systematics. 15:303-328
Okubo, A., J.D. Ackerman, and D P. Swaney. 2002. Passive Diffusion in Ecosystems. Pp.
31-106 in A. Okubo and S Levin (eds.) Diffusion and Ecological Problems: New
Perspectives, 2"^ ed. Springer Verlag. 467 pp.
Sand-Jensen, K. 2003. Drag and reconfiguration of freshwater macrophytes. Freshwater
Biology. 48:271-283.
Sand-Jensen, K., and O. Pedersen. 1999. Velocity gradients and turbulence around
macrophyte stand in streams. Freshwater Biology. 42:315-328.
Santamaria, L. 2002. Why are most aquatic plants widely distributed? Dispersal, clonal
growth and small-scale heterogeneity in a stressful environment. A cta O ecologica.
23:137-154.
Santamaria, L., J. Figuerola, J.J. Pilon, M. Mjelde, A.J. Green, T. De Boer, R.A. King,
and R.J. Gomall. 2003. Plant performace across latitude: the role of plasticity and
local adapatation in an aquatic plant. Ecology. 84(9):2454-2461.

Chapter 4

185

Schutten, J., and A.J. Davy. 2000. Predicting the hydraulic forces on submerged
macrophytes from current velocity, biomass and morphology. Oecologia. 123:445452.
Sculthorpe, C D . 1967. The Biology o f Aquatic Vascular Plants. Edward Arnold, London
(Reprinted 1985 Koeltz Scientific, Konigstein), 610 pp.
Sebens, K.P., J. Witting, and B. Helmuth. 1997. Effects of water flow and branch spacing
on particle capture by the reef coral Madracis mirabilis (Duchassaing and Michelotti).
Journal o f Experimental Marine Biology and Ecology. 211:1-28.
Sheath, R.G., and J.A. Hamhrook. 1988. Mechanical adaptations to flow in freshwater
red algae. Journal o f Phycology. 24:107-111.
Speck, O. 2003. Field measurements of wind speed and reconfigureation in Arundo
donax (Poaceae) with estimates of drag forces. American Journal o f Botany.
90(8): 1353-1256.
Stewart, H.L., and R.C. Carpenter. 2003. The effects of morphology and water flow on
photosynthesis of marine macroalgae. Ecology. 84(11):2999-3012.
Strand, J.A., and S.E.B. Weisner. 2001. Morphological plastic responses to water depth
and wave exposure in an aquatic plant (Myrophyllum spicatum). Journal o f Ecology.
89:166-175.
Schwinge, J., D.E., Wiley, and D.F. Fletcher. 2002. Simulation of the flow around spacer
filaments between narrow channel walls. 1. Hydrodynamics. Industrial and
Engineering Chemistry Research. 4 1 :2977-2987.
Vogel, S. 1994. Life in Moving Eluids The Physical Biology o f Flow. Princeton
University Press, Princeton, N.J. 467 pp.
Wells, C.L., and M. Pigliucci. 2000. Adaptive phenotypic plasticity: the case of
heterophylly in aquatic plants. Perspectives in Plant Ecology, Evolution and
Systematics. 3(1):1-18.

186

Table 4.1: ANCOVA results for (a) the area covered in dye and (b) the concentration of
dye in that area for the different domains for all of the test subjects combined. (Velocity =
average flume velocity; Plant = empty test section, circular cylinder, Elodea, Elodea with
the leaves at every second node removed, Elodea with the leaves at every second and
third node removed, Ceratophyllum, Ceratophyllum with the leaves at every second node
removed, and Ceratophyllum with the leaves at every second and third node removed;
Area = area of dye coverage; Cone. = concentration of dye; side/top = side and top
perspectives; before/after = upstream and downstream domain). Grey values are
significant.
(a)

Effect
Intercept
velocity
plant
side/top
before/after
plant*slde/top
plant*before/after
side/top*before/after
plant*side/top*before/after
Error
Total

(b )

Effect
Intercept
velocity
plant
side/top
before/after

Univariate Results for Each DV (ancova_spreadsheet_ch2)
Sigma-restricted param eterization
Effective hypothesis decomposition
Degr. of
Freedom
1
1
7
1
1
7
7
1
7
127
159

area
88

area
MS

area
F

area

213.8189
2.9844

213.8189
2.9844

380.6048
5.3124

0.000000
0.022797

18.0545
2.5104

2 5792
2.5104

0.000133
0 0::6478

526.8370
3.5427

526 8370

4.5911
4.4687
937 /Q72

13 6716
5.3689
3.6078
71.3470
647.9242]

1.9531
5. ..1689
0.5154

0.5061

P

0 00 00 00

0.9009
3 4766
9.5567
0.9174

0.508095
0 001927
0.002449
0.495361

0.5618
|

Univariate Results for Each DV (ancova_spreadsheet_ch2)
Sigm a-restricted param eterization
Effective hypothesis decom position
conc.
88

25.73763
5.30758
0.73668
0.36289

conc.
MS

conc.
F

conc.

25.73763

537.2375
110.7884

0.000000

5.30758
0.10524
0.36209

P

2 .1 0 6 /

0.000000
0.038608
0.006786
0.000703

0.57799

0.57799

7.5749
12.0648

plant*side/top
plant*before/after

0.11445

0.01635

0.3413

0.933496

0.62744

1.8710

side/top*before/after
plant*side/top*before/after.......
Error
T otal

0.07847
0.21367
6.08424

0.08963
0.07847
0.03052
0.04791 1

0.079586
0.202932
0.724427

14.103421

1

1.6380
0.6372

1

187

Table 4.2: ANCOVA results for the area of dye coverage and the concentration
of dye for (a) the circular cylinder, (b) Elodea, (c) Elodea with the leaves at
every second node removed, and (d) Elodea with the leaves at every second and
third node removed.
Univariate Resuits for Each DV (Cyiinder)
Sigma-restricted parameterization
Effective hypothesis decomposition
Degr. of
area
area
%
Freedom
Effect
F
interceat
1 22,0598 22.05984 22.21238
1
5.4365
5.47411
velocity
5.43651
1
0,4346
0.43756
0.43455
side.'toj)
before/after
1 2 ^ 6 8 3 9 95.68384 96.34554
0.33172
side/top'before/after___
1
0.3294
0.32944
Error
15 14.8970
0.99313
19 116.7814
Totai
(a )

(b )

Effect
intercept
velocity
side/top
before/after
s de/top‘before/after
Error
Totai

(c )

Effect
Intercept
velocity
side/top
before/after
side/top'before/after
Error
Totai

(d)
Effect
intercept
velocity
side/top
before/after
side/top'before/after
Error
Totai

conc.
conc.
area
conc.
conc.
F
88
P
P
0.000278 1.9786971 1.978697 1 95.67204 I 0.000000
0.033544 0.437853 1_0.437853 21.17066 ! 0.000346
0.518336 o . o i e s o T T 0.016504 1 0.79797 0.385807
1.61964 I 0.222516
0,000000 0.033497 0.033497
0.02070j 0.887524
0.573188 0.000428 0.000428
0.310231 0.020682
0.798514 !

Univariate Resuits for Each DV ( Elodea )
Sigma-restricted parameterization
Effective hypothesis decomposition
conc.
conc.
Degr. of
area
area
area
conc.
conc.
area
88
F
Freedom
88
MS
MS
P
1 40.63801 40.637981 111.66271 O.OOOOOdI 2.9495231 2.9495231 120.12691 0.000000
0.4366 0.51880611 0,6493321 0.6493321 26.4457) 0.0001%
1
0.1589 0.15888
1,2116 0.288375
.......... .......1
0 4885 0.48855 2 2 l .3424 0.264723 0.029748 0.029748
1.1508 0.300344
1 92.4244 92 4 2 ^ "253.9585 O.OOOOOa 6.028255 0.028255
3.21731 0.093045
1
1.7758 1.77681
4.8822 0.0431 o d 0.078995 0.078995
II 0.3683011 0.0245531
15
5.4590 0.36394
II 1.154631)
19 100.3077

Univariate Results for Each DV ( Elodea 2nd)
Sigma-restricted parameterization
Effective hypothesis decomposition
Degr. of
area
; area
' area
area
conc.
conc.
conc. , conc.
Freedom
88
'
MS
F
p
88
MS
F
i
p
l l 27.41857 27.418571 108.3929 O.OOOOOdI 1.351336 1.3513361 41.09747 0.000016
1 0.91044) 0.91044
3 5992 0.078628110.174365 0.174365' 5.30288 0.037153
1 0.07360 0.07360
0.2910 0.598070 0.116977 0.1169771 3.55756 0.Ô8Ô193
1 83.129361 83.12%6I 3 2 8 . ^ 3 O .O O O O od 0.108770 0.10877C 3.30796 0.090396
1 0.14255 0.14255
0.5635 0.465267 0.102537 0.102537 3 11839 0.099202
II 0 4 6 0 3 3 7 0.032881:
3.541381 0.25296
H 89.33794
I
II 0.999678)
1
iq

Univariate Results for Eacfi DV ( Elodea 3rd)
Sigma-restricted parameterization
Effective hypotfiesis decomposition_______
area
conc.
conc.
conc.
Degr. of
area
area
area
conc.
F
Freedom
88
F
88
MS
MS
_2_
28.67930i 28.679301 108.28451 O.OOOOOdI 3.083176) 3.0831761 96.756121 0.000000
1 0.020181
0.02018! 0.0762, 0.78626a 0.732808| 0.732808| 22.99695! 0.000236
1 0.38676
0.38676 1.4603 0 .245589 0.129212 Ô. 129212 4.054941 0.062343
1 56.47179i 56.47179i213.2207: O .O O pO od 0.049666 0.049666 1.55861 0.231004
0.08260 0.3119 0 .584784 0.0004771 0.000477 0.01498! 0.9Ô4222
1 0.08260!
151 3.97277:
0.26485:________ i _ _ _ _ _ II 0.4779821 0.031865!
1.390145
19 60.93410

188

Table 4.3; Post hoc comparisons between Elodea and its associated manipulations for the
area covered in dye in the (a) side perspective upstream, (b) in the top perspective
upstream, (c) in the side perspective downstream, and (d) in the top perspective
downstream. Significant results indicated in bold. {Elodea 2"“^ = Elodea with every
j

j

— —

Cylinder

--- ----- --------------------1

o rd _

Elodea

Elodea

A

* u :„ A

„ ^ A ^

Empty flume

Cylinder
Elodea

0.84

Elodea

0.94

0.9

EZofko 3"^

0.76

0.91

0.81

Empty flume

0.73

0 .8 8

0.79

Cylinder

Elodea

(b)

0.97
2 ""

Elodea 3'^'^ Empty flume

Cylinder
0.63

Elodea

0.91

0.71

EZotka 3""

0.93

0.70

0.98

Empty flume

0.99

0.62

0.90

Cylinder

Elodea

EZo(Zga 2 "''

(c)

0.92
Elodea 3*^^ Empty flume

Cylinder
0.72

Elodea
2 "''

0.24

0.13

EZofZga 3"^

0 .0 0 1

< 0 .0 0 1

0.039

Empty flume

0.003

0 .0 0 1

0.071

(d)

Cylinder

Elodea

EZojga 2 °"

0 .8

Elodea

Cylinder
Elodea

0.69

EZofZeo 2 "''

0.96

0.73

Elodea 3'^'^

0 .2 2

0.40

0.24

Empty flume

< 0 .0 0 1

< 0 .0 0 1

< 0 .0 0 1

< 0 .0 0 1

Empty flume

189

Table 4.4: ANCOVA results for the area of dye coverage and the concentration
of dye for (a) the circular cylinder, (b) Ceratophyllum, (c) Ceratophyllum with
the leaves at every second node removed, and (d) Ceratophyllum with the leaves
(a )

Effect
Interceat
velociiv
side/top
before/after
side/top'before/after
Error
Total

(b)

Effect
Intercept
velocity
side/top
before/after
side/top'before/after
Error
Total

(c)

Effect
intercept
velocity
side/top
before/after
side/top'before/after
Error
Totai

(d )

Effect
intercept
velocity
side/top
before/after
side/top'before/after
Error
Totai

Univariate Results for Each DV (Cyiinder)
Sigma-restricted parameterization
Effective hypothesis decomposition
Degr. of
area
area
area
Freedom
MS
SS
F
1 22.0598 22.05984 22.21238
1
5.43651
5.47411
5.4365
1
0.43455
0 43756
0.4346
1 95.6839 95.68394 96.34554
0,32944
1
0 3294
0.33172
0.99313
15 14.8970
19 116.7814

conc. 1 conc.
conc.
SS
MS
F
0.0002781 1.978697 i 1.978697 95.67204
0.0335441 0.437853 ■0.437853 21.17066
0.79797
0.518336 0.016504 0.016504
1.61964
0.000000 0.033497 0.033497
0 02070
0.573188 0.000428 0.000428
1 0.310231 0.020682
0.798514 1
i
area

conc.

D
0.000000
0.000346
0.385807
0 22251G
0.887524

Univariate Results for Each DV ( Ceratophyllum )
Sigma-restricted parameterization
Effective hypothesis decomposition
conc.
area
area
area
conc.
conc.
conc.
Degr. of
area
Freedom
MS
F
SS
MS
F
SS
P
P
1 43.71454 43.71454 109.4433 O.OOOOOdI 3.4922961 3.492296 32.72128 0.000041
__ 1 1.35136 1.35136 3.3833 0.0857381 0.446894 0.446894 4.18720 0.058675
1 1 69315 1.69315 4.2390 0 057310 0.152006 0.152006 1.42423 0.251243
1 65.79465 65.79465 164.7228 o.oooôôd Ô.005253! 0.005253 0.04921 0.827427
1 1.87377 2Î.8 7 37 7
4.6912 0.046839 0.033108 0.033108 2 0 .3 1 0 2 1 0.585767
15 5.99140 0.39943
1.6009291 0.106729
19 76.70433
!
II 2.2381901

Univariate Results for Each DV ( Ceratophyllum 2nd)
Sigma-restricted parameterization
Effective hypothesis decomposition
conc.
conc.
conc.
conc.
Degr. of
area
area
area
area
SS
MS
Freedom
SS
MS
F
P
P
1 20.26384 20.26384 65.0744 O.OOOOOll 2.4540231 2.454023 104.35301 0.000000
1 0.24563 0.24563
0.7888i 0.388484 0.291062 0.291062. 12.37691 0.003106
0.8197 0.379584 0.0610931 0.061093
2.59791 0.127844
1 0 25525 0.25525
6.3341 0.023707
1 49.58870 49.58870 159.2470 6.000000 0 1 4 8 9 5 7 0.148957
0.0819 0.778716
1 0.46875 0.46875
1.5053 0.238763 0.001925 0 001925
0.352748 0.023517
15 4.67092 0.31139
0.855785|
19 55.22925
i

Univariate Results for Each DV ( Ceratophyllum 3rd)
Sigma-restricted parameterization
Effective hypothesis decomposition
conc.
conc.
conc.
conc.
Degr. of
area
aioa
area
area
SS
F
Freedom
MS
F
MS
SS
P
P
1 21.88786 21.88786 28.40197 0.000084 4.9197711 4.9197711 106.5441 0.000000
1 2.13744 2.13744 2.77357 0.116574 1.097093 1.097093 23.7590 0.000202
t 0.11254 0.11254 0.14603 l_0.70771E 0.013358 0.013358
0.2893 0.598574
1 65.82507 65.82507 85.41545 o.oooooc 0.227015 0.227015
4.9163 0.042470
1.8466 0.194266
1 0.80929 0 80929 1.05014 0.321717 0.085267 0.085267
0.692639,1 0.046176
15 11.55969 0.77065
2.115372]
19 80.44402

190

Table 4.5: Post hoc comparisons between Ceratophyllum and its associated
manipulations for the area covered in dye in the (a) side perspective upstream, (b) in the
top perspective upstream, (c) in the side perspective downstream, and (d) in the top
perspective downstream. Significant results indicated in bold. {Ceratophyllum
=
G^stptophyllum with every second node removed; Ceratophyllum
= Ceratophyllum
with every seconc and thirc node removec I)
Ceratophyllum 2““* Ceratophyllum 3^‘* Empty flume
Cylinder Ceratophyllum
Cylinder
Ceratophyllum

0.91

Ceratophyllum

0.78

0.87

Ceratophyllum 3”^^

0 .8 8

0.98

0.89

Empty flume

0.73

0.82

0.95

Cylinder

Ceratophyllum

Ceratophyllum 2““*

(b)

0.84
Ceratophyllum 3^“* Empty flume

Cylinder
Ceratophyllum

0.98

Ceratophyllum 2"^

0.97

0.95

Ceratophyllum 3""

0.64

0 .6 6

0.62

Empty flume

0.99

0.97

0.98

Cylinder

Ceratophyllum

Ceratophyllum 2““*

(c)

0.63

Ceratophyllum 3^“* Empty flume

Cylinder
Ceratophyllum

0.35

Ceratophyllum 2““*

0.006

0.07

Ceratophyllum 3'^‘*

0.16

0.64

0.18

Empty flume

0,003

0.042

0.82

(d)

Cylinder

Ceratophyllum

Ceratophyllum 2'“'

0 .1 1

Ceratophyllum 3'“' Empty flume

Cylinder
Ceratophyllum

0.023

Ceratophyllum 2"^

0.007

0.67

Ceratophyllum 3'^‘*

0.16

0.38

0.19

< 0.001

0.005

0.015

Empty flume

< 0.001

191

Table 4.6: Post hoc comparisons between Elodea and its associated manipulations for the
concentration of dye in the (a) side perspective upstream, (b) in the top perspective
upstream, (c) in the side perspective downstream, and (d) in the top perspective
downstream. Significant results inc icated in bold.
Elodea 3"^^ Empty flume
EW ga
(a)
Cylinder Elodea
Cylinder
Elodea

0.74

Ekdga 2""

0.54

0.34

Elodea 3"^^

0.58

0.38

0.95

Empty flume

0.70

0.47

0.82

Cylinder

Elodea

Moffgo 2°"

(b)

0.87
Elodea 3*^“^ Empty flume

Cylinder
Elodea

0.72

Ekdga 2""

0.37

0.59

0.74

0.49

0.22

0.89

0.82

0.45

Cylinder

Elodea

EZodgo 2"^

Empty flume
(c)

0.64
Elodea 3"^*^ Empty flume

Cylinder
Elodea

0.53

kodga 2""

0.56

0.96

EZodgo 3"^

0.37

0.79

0.75

Empty flume

0.002

0.012

0.011

(d)

Cylinder

Elodea

EZodgo 2"^

0.024
Elodea 3^^ Empty flume

Cylinder
Elodea

0 .1 1

EZodga 2"'^

0.075

0.85

EZodga 3"^

0.13

0.91

0.77

Empty flume

0.029

0.56

0.69

0.49

192

Table 4.7: Post hoc comparisons between Ceratophyllum and its associated
manipulations for the concentration of dye in the (a) side perspective upstream, (b) in the
top perspective upstream, (c) in the side perspective downstream, and (d) in the top
perspective downstream . Significant resu ts indicated in bo d.
(a)

Cylinder

Ceratophyllum

Ceratophyllum 2"^

Ceratophyllum 3"^^ Empty flume

Cylinder
Ceratophyllum

0.33

Ceratophyllum 2”‘*

0.65

0.15

Ceratophyllum 3*^^

0.42

0.87

0.20

Empty flume

0.70

0.17

0.95

Cylinder

Ceratophyllum

Ceratophyllum 2““*

(b)

0.23
Ceratophyllum 3'^ Empty flume

Cylinder
Ceratophyllum

0.19

Ceratophyllum 2“‘*

0.89

0.24

Ceratophyllum 3^“*

0.92

0.16

0.81

Empty flume

0.89

0.15

0.79

Cylinder

Ceratophyllum

Ceratophyllum 2"^

(c)

0.97

Ceratophyllum 3"* Empty flume

Cylinder
Ceratophyllum

0.42

Ceratophyllum 2”“*

0.11

0.44

Ceratophyllum 3'“*

0.040

0.21

0.63

Empty flume

0.002

0.020

0.11

(d)

Cylinder

Ceratophyllum

Ceratophyllum 2”‘*

0.27
Ceratophyllum 3^“* Empty flume

Cylinder
Ceratophyllum

0.031

Ceratophyllum 2”**

0.080

0 .6 8

Ceratophyllum 3'^'*

0.004

0.47

0.26

Empty flume

0.029

0.98

0.66

0.48

193

Table 4.8: (a) ANCOVA and (b) post hoc comparisons for the angle of the
Ceratophyllum leaves (and the manipulations thereof) from the stem in the up and
downstream position. Significant results indicated in bold.

(a)

Effect
Intercept
velocity
plant
fro n t/b a c k
p l a n ; 'f r o n t / b a c k

Error
Total

Univariate Results for Each DV (leaf angle)
Sigma-restricted parameterization
Effective hypothesis decomposition
Degr. of
angle
angle
angle i angle
Freedom
88
MS
F
1
P
1 167470.3 167470.3 845.3146 0.000000
1
1556.2
7.8551 0 005646
1556.2
5.4250 0.005187
2
2149.5
1074.8
1
436.3
436.3
2.2022 0.139636
2
3.2333 0.041823
640.6
1281.1
173 34274.0
198.1
179 39944.6
1
1

Cento 0^ Cento limn Cento 2ndl%) Cento 2ndDa\ui Cento 3idlJ) Cento 3rd Ii)«ii

(b )

Cemto Op
Cento D o v h i

010019

Cento 2ndOp

aoo©

0.93

Cento 2nd Down

aoosi

0.86

0.94

Cento SdOp

<aooi

0.33

0.31

0.27

Cento ^ D d\ui

<aooi

0.43

0.40

0.36

0.86

194

Table 4.9: (a) A NCO V A and (b) post hoc com parisons for the angle o f deflection
between E lodea and Ceratophyllum and there associated manipulations. Significant
results indicated in bold.

Effect
Intercept
Velocity
Plant
Error
Total

Univariate Results for Each DV (plant angle)
Sigma-restricted parameterization
Effective hypothesis decomposition
Degr. of
Angle ■ Angle
Angle
Angle
Freedom
88
!
MS
.
F
p
81.181
81.181
1.113041 0.298862
1
1 7072.331 7072.327 96.96575 0.000000
6 7655.67 1275.944 17.49395 0.000000
34 2479.84
72.9361
1
41 17207.831
I

(b)

cylinder

(a)

Elodea

Elodea ^id

Elodea 3rd

Cerato

Cerato 2nd

Qlinder
Elodea

0.0018

Elodea 2nd

0.53

0.0097

Elodm 3rd

0 .M8

0.19

0.16

Cerato

<0.001

<0.001

<aooi

<0.001

Cerate 2nd

<0.001

0.70

0.003

0.098

<aooi

Cerato 3rd

<0.001

0.10

<0.001

O.OOS3

0.0031

0.21

Cerato 3rd

195

Table 4.10: (a) ANCOVA and (b) post hoc comparisons for the frequency of oscillations
for Elodea, Ceratophyllum, and there associated manipulations. Significant results
indicated in bold.
Univariate Results for Eacfi DV (oscillations)
Sigma-restricted parameterization
Effective hypothesis decomposition

(a)

Degr. of
Freedom
Effect
Intercept
Velocity
Plant
Error
Total

6
27
34

(b)

Cylinder
Cylinder
Elodea
Elodea 2nd
Elodea 3rd
Cerato
Cerato 2nd
Cerato &id

oscillations per . oscillations per oscillations per
second
second
second
MS
F
88
P
0.39643
0.534231
0.007968
0.007968
57.479m
0.000000
1.155310
1,155310
4.78538
0.001925
0 .5 7 /; 07
0 .0 9 6 '8f:
0.020100
0.542690

oscillations per
second
1

1

<0.001
0.005

2.275107}

Elodea

< aooi

0.14
0.61

<0.001

0.83

0.001

0.31
0.59

<0.001

1

Elodm 2nd Elodm 3rd

1

Cerato

Cerato 2nd

0.42
0.74

0.63

032
0.20
0.63
0.34

0.76
0.60
0.97

Cerato S d

196

Figure 4.1: Flow visualization around a 0.7 cm diameter circular cylinder using
fluoroscein dye in the side (left) and the top (right) perspectives. Velocities increase
from top to bottom panels [(a) and (b) are 1.3 cm s'\ (c) and (d) are 2.0 cms"\ (e) and (f)
are 5.0 cms"\ (g) and (h) are 8.4 cm s'\ and (i) and (j) are 11.0 cms'*]. Flow is from left
to right. Top and side photos were taken sequentially.

197

Side
0.0

1.5

Top

Reynolds number (x 10 )
3.0

4.5

6.0

7.5

0.0

1.5

3.0

4.5

6.0

7.5

10

(a ) y = (0.01110.006)X + 0.43 1 0 .0 4
F = 0.47, p = 0.19

8

i

y = (0.003±0.002)X + 0.410.1
F = 0.36, p = 0.28

(b )

6

I

4

'B'

î

:
10

y = (0.49± 0.10)x + 2.410.7

I

y = (0 .0 5 1 0 .0 5 )x + 4.310.3
= 0.22, p = 0.41

f = 0.88, p = 0.01^

8
6

I

4

Q

2
0

2.0

(e)

y = (-0 .0 5 1 0 .0 2 )x + 0.610.1

y = (-0 .0 3 i0 .0 2 )x + 0 .6 1 0 .2
= 0.31, p = 0.32

(g )

y = (-0.0510.01)x + 0.54+ 0.06

y = (-0 .0310.02)x + 0.510.1
f = 0.47, p = 0.19

W)
0.5

0.0
2.0

0.5
0.0
0

2

4

6

8

10

12

0

2

4

6

8

10

12

Velocity (cms'^)
Figure 4.2: Regression of area of dye coverage (a-d) and concentration of dye (e-h)
versus velocity or Reynolds number (based on the diameter) in the different domains for
the circular cylinder: (a) and (e) upstream in the side perspective, (b) and (f) upstream in
the top perspective, (c) and (g) downstream in the side perspective, and (d) and (h)
downstream in the top perspective. Solid line is the regression and dotted lines are the
95% confidence limits of the mean.

198

Figure 4.3: Flow visualization around Elodea canadensis using fluoroscein dye in the
side (left) and the top (right) perspectives. Velocities increase from top to bottom panels
[(a) and (b) are 1.3 cm s'\ (c) and (d) are 2.0 cms"\ (e) and (f) are 5.0 cm s'\ (g) and (h)
are 8.4 cms ', and (i) and (j) are 11.0 cms'^]. Flow is from left to right. Top and side
photos were taken sequentially.

199

Top

Side
Reynolds number (x 10 )
8

10

12

8

10

12

10

(a )
8

y = (-o.oi±o.oi)x + o.46±o.07
_

mo

r, —

n "Î/1

(I)) y = (0.009±0.010)x + 0.63±0.06
= 0.21, p = 0.43

6

4
2

I

0

a, 1 0
‘o

II

(c ) y = (-0.0004±0.0004)x + 5.3±1.0

f < 0 .001 , p = 0r99

8

y = (0.10±0.02)x + 3.8±0.2
= 0.84, p = 0.026

6

c/3

I

Q

4
2
0

(-0.04±0.02)x + 0.6±0.1

y = (-0.03±0.02)x + 0.5±0.1
= 0.44, p = 0.21

y = (-0.06±0.02)x + 0.7±0.1
= 0.77, p = 0.049

y = (-0.06±0.02)x + 0.9±0.1
= 0.78, p = 0.046

I
g
§
a

2.0

1.5
§

(g )

O

I Q 1.0

I

Q

Velocity (cms'^)
Figure 4.4; Regression of area of dye coverage (a-d) and concentration of dye (e-h)
versus velocity or stem Reynolds number (based on the plant diameter) in the different
domains for Elodea canadensis: (a) and (e) upstream in the side perspective, (b) and (f)
upstream in the top perspective, (c) and (g) downstream in the side perspective, and (d)
and (h) downstream in the top perspective. Solid line is the regression and dotted lines are
the 95% confidence limits of the mean.

200

Figure 4.5: Flow visualization around Elodea with the whorl at every second node
removed using fluoroscein dye in the side (left) and the top (right) perspectives.
Velocities increase from top to bottom panels [(a) and (b) are 1.3 cms"\ (c) and (d) are
2.0 cms"\ (e) and (f) are 5.0 cms ’, (g) and (h) are 8.4 cm s'\ and (i) and (j) are 11.0 cms‘
*]. Flow is from left to right. Top and side photos we re taken sequentially.

201

Top

Side
Reynolds number (x 10 )
8

10

12

0

^ 2 _ n 0*7

« — n '3 '7

10

12

y = (0. l± 0 .1 )x + 4.0+0.7
= 0.23, p = 0.41

y = (0.21± 0.04)x + 3.410.3
? = 0.89, p = 0.015

8

8

4

(b) y = (-a.007±0.007)x + 0.54+0.05
^ = 0.22, p = 0.42

y = (o.oi±o.oi)x + o.39±o.08

(a)

2

6

4
2
0
2 .0

fe)

y = (-o.03±o.02)x + o.4±o. i

(f)

y = (0.03± 0.01)x + 0.14+ 0.04
..2 _ n

M

0 .0

y = (-0.06+ 0.01 )x + 0.91+ 0.08
0.88, p 0.016

y = (-0.06+ 0.02)x + 0.7+0.1

0.68, p = 0.082

I

CO

I

Q

0 .0

'

' -■-■
■—
10

12

0

2

Velocity (cms'*)
Figure 4.6: Regression of area of dye coverage (a-d) and concentration of dye (e-h)
versus velocity or stem Reynolds number (based on the plant diameter) in the different
domains for Elodea canadensis with every second node removed: (a) and (e) upstream in
the side perspective, (b) and (f) upstream in the top perspective, (c) and (g) downstream
in the side perspective, and (d) and (h) downstream in the top perspective. Solid line is
the regression and dotted lines are the 95% confidence limits of the mean.

202

Figure 4.7; Flow visualization around Elodea with the whorls at every second and third
node removed using fluoroscein dye in the side (left) and the top (right) perspectives.
Velocities increase from top to bottom panels [(a) and (b) are 1.3 cm s'\ (c) and (d) are
2.0 cms"\ (e) and (f) are 5.0 cm s'\ (g) and (h) are 8.4 cm s'\ and (i) and (j) are 11.0 cms"
’]. Flow is from left to right. Top and side photos were taken sequentially.

203

Top

Side
Reynolds number (x 10 )
2

4

8

10

12

0

(a) y = (0.002±0.01 l) x + 0.33±0.07

2

4

8

10

12

r y ) y = (-0.02+ 0.01)x + 0.58±0.07
,2_^jj„_n--

,•2 _ n nr « - n

I

I

8

i
sc/3

6

O

2

I

y = (-0.10± 0.06)x + 4.5+ 0.4
i = 0.45, p = 0.20

y = (0 .08± 0.10)x + 3.2±0.7
= 0.16, p = 0.50

4

0
2 .0

(e)

y = (-0.06+ 0.04)x + 0.6±0.2

(f)

y = (-0.03+ 0.01)x + 0.60+ 0.06
^2 _ n "7/1 » - n AAn

(h)

y = (-0.05± 0.01)x + 0.82+ 0.05

= 0.44, p = 0.22

I

W)
0.5
0 .0
2 .0

(a)

y = (-0.07± 0.02)x + 0.8±0.I
,2_nci «-nn-ie

B
c3
0.5
0 .0
0

2

4

6

8

10

12

0

2

8

10

12

Velocity (cms"^)
Figure 4.8: Regression of area of dye coverage (a-d) and concentration of dye (e-h)
versus velocity or stem Reynolds number (based on the plant diameter) in the different
domains for Elodea canadensis with every second and third node removed: (a) and (e)
upstream in the side perspective, (b) and (f) upstream in the top perspective, (c) and (g)
downstream in the side perspective, and (d) and (h) downstream in the top perspective.
Solid line is the regression and dotted lines are the 95% confidence limits of the mean.

204

Figure 4.9: Flow visualization around Ceratophyllum demersum using fluoroscein dye in
the side (left) and the top (right) perspectives. Velocities increase from top to bottom
panels [(a) and (b) are 1.3 cm s'\ (c) and (d) are 2.0 cm s'\ (e) and (f) are 5.0 cms"\ (g)
and (h) are 8.4 cm s'\ and (i) and (j) are 11.0 cms ']. Flow is from left to right. Top and
side photos were taken sequentially.

205

Top

Reynolds number (x 10^)
0 .0

10

4 ,5

(a)

9 .0

1 3 .5 1 8 .0 2 2 .5 2 7 .0

y = (0.02± 0.01)x + 0.31±0.07
= 0.56, p = 0.14

8

0 .0

4 .5

(b)

9 .0

1 3 .5 1 8 .0 2 2 .5 2 7 .0

y = (-0 .01± 0.02)x + 0.5±0.1
= 0.14, p = 0.51

6

I

4

S'
2

I

i
§

I

I

o
P

0
10
y = (-0.2± 0.1)x + 5.6±0.9
= 0.32, p = 0.31

y = (-0.12± 0.03)x + 4.2± 0.2
= 0.84, p = 0.028

/g x

y = (-0.001± 0.062)x + 0.5+0.4
r^ = 0.00051, p = fir9 8

y = (-0.07± 0.04)x + 0.9± 0.2
= 0.52, p = 0.16

(g )

y = (-0.07± 0.04)x + 0.7±0.2
= 0.53, p = 0.16

y = (-0.03± 0.02)x + 0.8±0.1
= 0.48, p = 0.19

8
6

4
2
0
2 .0

1.5
1 .0

I

0.5
0 .0

g

I

2 .0

1.5
1

I O§

1 .0

I Q 0.5

o
Q

0 .0

0

2

4

6

8

10

12

Velocity (cms' )
Figure 4.10: Regression of area of dye coverage (a-d) and concentration of dye (e-h)
versus velocity or stem Reynolds number (based on the diameter) in the different
domains for Ceratophyllum demersum: (a) and (e) upstream in the side perspective, (b)
and (f) upstream in the top perspective, (c) and (g) downstream in the side perspective,
and (d) and (h) downstream in the top perspective. Solid line is the regression and dotted
lines are the 95% confidence limits of the mean.

206

• y = (-0.07±0.32)x + 14±2, r = 0.01, p 0.84
o y = (-1.3±0.4)x + 49±2, = 0.74, p = 0.027

• y = (-l± l)x + 74±1, = 0.87, p = 0.006
o y = (-2.2±0.4)x + 43±2, = 0.88, p = 0.006
100

80
60
40
20
0

• y = (-1.0±0.3)x + 51±2,
o y = (-1.2±0.3)x + 56±2,

• y = (-1.6±0.7)x + 40±4, / = 0.55, p = 0.09
y = (-0.4±0.2)x + 50+1, = 0.45, p = 0.14

= 0.77, p = 0.02
= 0.77, p = 0.02

100

I
I

t

80
60
40
20
0

J
• y = (-2.3±0.9)x + 38±5,
o y = (-0.3±0.3)x + 55+2,

y = (-0.33±0.31)x + 16±2, = 0.22, p = 0.34
y = (l.l±0.5)x + 33±3, = 0.50, p = 0.11

= 0.62, p = 0.06
= 0.20, p = 0.36

100

80
60
40
20
0
0

2

4

6

8

10

12

0

2

4

6

8

10

12

Velocity (cms'^)
Figure 4.11: Regression analysis for the leaf angle from the stem of Ceratophyllum in the
upstream (i.e., faces the oncoming flow; • ) and downstream (i.e., in the wake of the flow;
o) position from the plant stem (a) is the bottom most leaf, (b) is the second leaf from the
bottom, (c) is the third leaf from the bottom, (d) is the fourth leaf from the bottom, (e) is
the fifth leaf from the bottom, and (f) is the top most leaf (apex).

207

Figure 4.12: Flow visualization around Ceratophyllum with the whorl at every second
node removed using fluoroscein dye in the side (left) and the top (right) perspectives.
Velocities increase from top to bottom panels [(a) and (b) are 1.3 cms \ (c) and (d) are
2.0 cms"', (e) and (f) are 5.0 cm s'\ (g) and (h) are 8.4 cm s'\ and (i) and (j) are 11.0 cms'
^]. Flow is from left to right. Top and side photos were taken sequentially.

208

Side
1.0

4 .5

Top

Reynolds number (x 10 )

9 .0

1 3 .5 1 8 .0 2 2 .5 2 7 .0

0 .0

4 .5

(b)

(a ) y = (o.oi5±o.oo4)x + 0 .27 +0.03
.-2 _ f\ on T. _ n

1 3 .5 1 8 .0 2 2 .5 2 7 .0

y = (0.02± 0.01)x + 0.32+ 0.06
= 0.62, p = O. l l

y = (0.09± 0.08)x + 2 .810.6
? = 0.29, p = 0.34

f = 0.0023, p = 0.93

8

9 .0

6

4
2
0
2 .0

(f)

y = (-0.02± 0.02)x + 0.4±0.1
f = 0.21, p = 0.43

y = (-0.01+ 0.02)x + 0.510.1
= 0.16, p = 0.50

0.5
0 .0
2 .0

( h ) y = (-0.03110.005)x + 0.7510.03

§

^2 _ AO': « - n nm

f = 0.80, p = 0.039

I U§

I I

I

Q

0.5
0 .0
0

2

4

6

8

10

12

0

2

4

6

8

10

12

Velocity (cms ')
Figure 4.13: Regression of area of dye coverage (a-d) and concentration of dye (e-h)
versus velocity or stem Reynolds number (based on the diameter) in the different
domains for Ceratophyllum demersum with every second node removed: (a) and (e)
upstream in the side perspective, (h) and (f) upstream in the top perspective, (c) and (g)
downstream in the side perspective, and (d) and (h) downstream in the top perspective.
Solid line is the regression and dotted lines are the 95% confidence limits of the mean.

209

• y = (0.46±0.03)x + 53±2, = 0.32, p = 0.23
o y = (-0.1±0.9)x + 52±6, ? = 0.003, p = 0.91

y = C-1.8±0.5)x + 48±3,
y = (-1.0±1.2)x + 37±8,

= 0.78, p = 0.019
= 0.14, p = 0.45

+ 54±2, = 0.31, p = 0.24■
- y =v(-0.5+0.3)x
-------------------------------------------------o y = (0.5+0.l)x + 53+1, = 0.70, p = 0.036

• y = (-1.6±0.2)x + 43.7±0.9, r" = 0.96, p < 0.001
o y = (-0.7+1.6)x + 31+10, = 0.044, p = 0.68

100

I
I0
t
1

80
60
40
20
0

• y = (-3.2±0.6)x + 52±3, = 0.89, p = 0.005
o y = (0.6±0.4)x + 48±3, = 0.36, p = 0.02
100
80
60
40

20
0
0

2

4

6

8

10

12

Velocity (cms'^)
Figure 4.14; Regression analysis for the leaf angle from the stem of the Ceratophyllum
with every second node removed in the upstream (i.e., faces the oncoming flow; • ) and
downstream (i.e., in the wake of the flow; o) position from the plant stem (a) is the
bottom most leaf, (b) is the second leaf from the bottom, (c) is the third leaf from the
bottom, (d) is the fourth leaf from the bottom, and (e) is the top most leaf (apex).

210

Figure 4.15: Flow visualization around Ceratophyllum with the whorl at every second
and third node removed using fluoroscein dye in the side (left) and the top (right)
perspectives. Velocities increase from top to bottom panels [(a) and (b) are 1.3 cm s'\ (c)
and (d) are 2.0 cm s'\ (e) and (f) are 5.0 cm s'\ (g) and (h) are 8.4 cm s'\ and (i) and (j)
are 11.0 cms'^]. Flow is from left to right. Top and side photos were taken sequentially.

211

Reynolds number (x 10 )

Side

0.0 4.5 9.0 13.5 18.0 22.5 27.0

10

0 .0

4 .5

Top
9 .0

1 3 .5 1 8 .0 2 2 .5 2 7 .0

(a) y = (0.010±0.005)X + 0.36±0.04

(b) y = (0,023± 0.006)x + 0.54+ 0.04

= 0.57, p = 0.14

= 0.82, p = 0.033

8
6

I

4

S'
2

1

0

I

10

"a
"o

(d)

y = ( 0 .2 4 ± 0 .2 0 ) x ^ 3 ± l
r^ = 0 .3 > ^ = 0.31

(C )

8

y = (0.07± 0.04)x + 3.5+0.2
f = 0.54, p = 0.15

I(/) I 4
6

I

Q

2
0

0

2

y = (-0.06± 0.02)x + 0.8±0.1
= 0.77, p = 0.051

f = 0.72, p = 0.069

y = (-0.08+ 0.04)x + 1.0+0.3
= 0.59, p = 0.12

y = (- 0 .0 7 + 0 .0 2 ) x + 1.1±0.2
= 0.74, p = 0.061

4

6

8

10

12

0

2

4

6

8

10

12

Velocity (cms'^)
Figure 4.16: Regression of area of dye coverage (a-d) and concentration of dye (e-h)
versus velocity or stem Reynolds number (based on the diameter) in the different
domains for Ceratophyllum demersum with every second and third node removed: (a)
and (e) upstream in the side perspective, (b) and (f) upstream in the top perspective, (c)
and (g) downstream in the side perspective, and (d) and (h) downstream in the top
perspective. Solid line is the regression and dotted lines are the 95% confidence limits of
the mean.

212

• y = (-1.3±0.4)x + 61±2, = 0.72, p = 0.031
o y = (0.4±0.5)x + 54±3, = 0.12, p == 0.48

• y = (-0.05±0.46)x + 58±3, = 0.003, p = 0.9
o y = (0.4±0.8)x + 37±5 , i = 0 .0 5 , p = 0.66
100

80

(a )

60

40
20
0

• y = (-1.8±0.5)x + 59±3, = 0.74, p = 0.026
o y = (l.l± 0 .5 )x + 57±3, = 0.59, p = 0.07

• y: =(-2.0±0.2)x + 38+1,
o y : (-0.7±l)x + 35±6,

= 0.95, p < 0.001
= 0.12, p = 0.50

100

80
60
40

20
0
0

2

4

6

8

10

12

0

2

4

6

8

10

12

Velocity (cms‘^)
Figure 4.17: Regression analysis for the leaf angle from the stem of the Ceratophyllum
with every second and third node removed in the upstream (i.e., faces the oncoming flow;
•) and downstream (i.e., in the wake of the flow; o) position from the plant stem (a) is
the bottom most leaf, (b) is the second leaf from the bottom, (c) is the third leaf from the
bottom, and (d) is the top most leaf (apex).

213

(b) downstream leaves

(a) upstream leaves

§
’a

I
"o

II
I
§
o
CL
%

3

-

2 - 1 0

1

2

Rate of change (7cms~ )
Ceratophyllum
Ceratophyllum w ith the w horls at every second n o d e rem o v ed
Ceratophyllum w ith th e w horls at every second and th ird n o d e rem o v ed
Figure 4.18: The rate of change in the angle of the leaf from the stem with change in
velocity for Ceratophyllum and its associated manipulations for the leaves (a) facing the
approaching flow (i.e., on the upstream side of the plant) and (b) in the downstream flow
(i.e., on the downstream side of the plant).
80
60
40
20
0

-20

(a)

y

= (0 ± 0 )x + 0 + 0
p = -

214

80

y = (3.1±0.3)x + 3±2
= 0.96, p < 0.001

60
40

O
y = (5.2±0.8)x+ 19±5
= 0.91, p = 0.003

§

o

y = (3.1±0.3)x - 11+2
r = 0.95, p < 0.001

<

y = (4.4±0.5)x - 0.6±2.9
= 0.95, p < 0.001

80
60

(d)

y = (1.9±0.2)x + 2±l
= 0.96, p < 0.001

(g)

40

20

"

0

y = (5.1±0.5)x4-3±3
= 0.96, p < 0.001

-20
8

10

8

12

10

12

Velocity (cms'^)
Figure 4.19: Regression results for the angle of deflection versus velocity for (a) the
circular cylinder, (b) Elodea, (c) Elodea with every second node removed, (d) Elodea
with every second and third node removed, (e) Ceratophyllum, (f) Ceratophyllum with
every second node removed, and (g) Ceratophyllum with every second and third node
removed. Solid line is the regression and dotted lines are the 95% confidence limits of the
mean.
0.8
y = (0 ± 0 )x + 0 ± 0

0.6
0.4

0.2
0.0

= -, p = -

215

0 .8

(b)
0.6

y = (0.06+0.01)x t//0.04±O

y = (0.05±0.023i + 0.1±0.1
0 .6 8 fp

0.4
0.2
0.0

0.8

(c) y = (0.05±0.01)x - 0.03±0.09
= 0 .8 W = 0.031

0.6

§

(f) y = (0.07±0.01)x - 0.07±0.p7^
— r\ ciA « _ f i'tin /:

0.4
0.2

O
c/5
O 0.0
'o

I

(d) y = (0.05+0.02)X/+ 0.07±0.14

"=

0

= 0 .0 9 7 ^ /

6

8

10

12

(g)

y = (0.05±0.02)xy2.1±0.1

2

4

6

8

10

12

Figure 4.20: Regression results for the frequency of oscillation with increasing velocity
for (a) the circular cylinder, (b) Elodea, (c) Elodea with every second node removed, (d)
Elodea with every second and third node removed, (e) Ceratophyllum, (f) Ceratophyllum
with every second node removed, and (g) Ceratophyllum with every second and third
node removed. Solid line is the regression and dotted lines are the 95% confidence limits
of the mean.

216

12
10

8

<0

I

6

y
I

4

2
0
Cylinder

E

E 2nd

E 3rd

0

0 2nd C 3rd

0 .8

(b)

05

E

0 .6

o

X

CM

F
3.

0.4

0)
05
c
(0
.c
o

0 .2

T

o

0 .0

f“
CC

tr

- 0 . 2

Cylinder

E

E 2ndE

3rd

C

C 2ndC

3rd

Cylinder

E

E 2 nd E 3rd

C

C 2 nd C 3rd

0.00
Ü

-0.02

o? - 0 . 0 4
-0.06
o

-0.08
-

0.1 0

Figure 4.21: Comparison of the fluid dynamic conditions around plants and a physical
model (a) The velocity at which the transition from laminar to turbulent occurs for the
different plants; (b) Rate of change in the area of dye coverage with change in velocity
for the different test subjects; and (c) The rate of change in the concentration of dye with
change in velocity for the different test subjects. The dotted line shows the comparison
with respect to the physical model. (E = Elodea', E 2°^ = Elodea with the whorl at every
second node removed; E 3'^'^ = Elodea with the whorls at every second and third node
removed; C = Ceratophyllum; C 2"“^= Ceratophyllum with the whorl at every second
node removed; and C
= Ceratophyllum with the whorl at every second and third node
removed)

217

CO
b
Ü
0)

7

T

6

-

5

-

4

-

3

-

2

-

1

-

0

-

O )
c
(U
O )
c
(Q
.c
Ü
o
CD
■+—'
(0
X

Cylinder Elodea E 2nd

E 3rd

Cerate C 2nd

C 3rd

E 3rd Cerato C 2nd

C 3rd

0.10
«
E 0.08 X
0.06 0)
O)
c
CO
o

0.04

O
O 0.02
B
CO

DC

0.00
Cylinder Elodea E 2nd

Figure 4.22: (a) Rate of change in the angle of deflection with change in velocity for the
different test subjects; and (b) the rate of change in the number of oscillations per second
with change in velocity for the different test subjects. The cylinder is always at zero
because it was rigid and as such remained stationary.

218

Chapter 5: Conclusion to Thesis
Aquatic plants are important constituents of freshwater systems because they act
as ecosystem engineers, in that they alter the physical environment, which leads to a wide
range of other changes (Wright et al., 2004). Of these physical factors, the flow
environment has the potential to limit macrophyte productivity at both low and high
velocities. At low velocity, the potential for diffusional stress is high due to the
existence of a thick boundary layer (Hurd et al., 1997; Schutten and Davy, 2000),
whereas at high velocities, the potential for damage or breakage is dominant due to the
increase in drag forces acting on the plant (Sand-Jensen, 2003). Macrophytes can be
plastic in their morphology, exhibiting heterophylly and foliar plasticity with changing
environmental conditions. It is reasonable to suggest that fluid dynamics may have
played a determining role in plant architecture.
This thesis investigated the interactions between a number of characteristically
shaped freshwater macrophytes and their flow environment. A review of the existing
data on plant-flow interactions was presented in Chapter 2. Chapter 3 examined the
interaction of linear-bladed, whorled and dissected leaf plants with flow by examining
flow patterns qualitatively and quantitatively. Chapter 4 examined the effect of the
plastic trait of intemodal spacing on macrophyte-flow interaction in whorled and
dissected leafed macrophytes using the same methodology to macrophytes that were
manipulated experimentally.
Experiments revealed that the linear-bladed species Vallisneria, and the whorled
species Elodea, affected the flow similar to a circular cylinder (used as a physical model).
Conversely, the whorled dissected leafed species, Ceratophyllum acted more like a mesh.

219

in that it broke the flow up into smaller eddies. In all cases the transition from laminar to
turbulent flow occurred at slower velocities for plants than the cylinder indicating an
increase in local mixing. This would be beneficial to the plants as it would replenish the
nutrients near the leaf surface, and thus promote the flux of nutrients and dissolved gases
across the boundary layer to the leaf surface (Anderson and Charters, 1982; Hurd et al.,
1996), thereby increasing productivity (Borchardt, 1994). Another important difference
between the different plant morphologies and the circular cylinder was that both the rate
of change in area and concentration of the dye were slower for plants than for the
cylinder. This indicates that plants retain fluid for a longer time than a cylinder, perhaps
providing more opportunity for uptake. In terms of plant compliance, the rate of change
in angle of deflection was smallest for Vallisneria (linear-bladed) and greatest for
Ceratophyllum (dissected), indicating that Vallisneria interacts less with the flow.
In terms of the effects of manipulated intemodal spacing on the flow (Chapter 4),
the transition from laminar to turbulent flow occurred at faster velocities for plants with
increased intemodal spacing than for the cylinder. This suggests a decrease in local
mixing, which indicates that increased intemodal spacing is a response to slower flows,
rather than to other factors (e.g., nutrient flux; Idestam-Almquist and Kautsky, 1995).
However, plants with increased intemodal spacing are generally exposed to higher flow
conditions because they exist higher in the water column, generally to increase light
interception (Ennos, 1999; Santamaria, 2002; Cronin and Lodge, 2003). The rate of
change in area of dye coverage decreased at a slower rate for the plants than for the
eylinder, as did the rate of change in the concentration of dye, reiterating the observation
that plants retain more fluid than a physical model. Whereas this indicates an increase in

220

the opportunity for nutrient uptake by the plants, it should be noted that the fluid is not as
well mixed as the fluid around intact plants and is retained for a longer period of time.
Elodea with the whorl at every second and third node removed, interacted the least of the
plants of this species with the flow, demonstrating that intemodal spacing does have an
effect on the downstream fluid patterns. All of the Ceratophyllum plants exhibited a high
plant-flow interaction, but the unmanipulated plant was the most compliant,
demonstrating that intemodal spacing has an effect on how the plants responded to the
flow.
This thesis showed that freshwater macrophytes of different morphologies have
different ways of interacting with the flow, while retaining the same amount of fluid.
Vallisneria may be better suited to faster flow environments because it had the least flow
interaction shown by the angle of deflection in the flow compared to the whorled and
dissected morphologies. This would indicate that drag forces would be less for this
linear-bladed architecture (Sheath and Hambrook, 1988; Gutierrez and Femandez, 1992;
Schutten and Davy, 2000). This is consistent with observations that smooth bladed
morphologies in macroalgae are found in faster flow environments than other
morphologies (Ackerman and Okubo, 1993; Hurd and Stevens, 1997). Conversely, it is
expected that Ceratophyllum would be better suited to slower flow environments because
it had the greatest flow interaction that led to the creation of small scale mixing, which
would enhance nutrient flux (Anderson and Charters, 1982; Nowell and Jumars, 1984;
Gutschick, 1999) and thus productivity. It is likely that Elodea is best suited to a flow
environment between these two extremes because, at low flows the plant-flow interaction
generated a large area of recirculation and many filamentous loops of attached dye

221
whereas at higher flows it deflected in the flow and generated vortices. In terms of
phenotypic plasticity, increased intemodal spacing was found to decrease the plant-flow
interaction in Elodea, as is evident in the reduced recirculation of dye, and in
Ceratophyllum, as evident from a decrease in the degree of compliance. An increase in
interhodal spacing would be beneficial in relatively low flow environments because there
would be less drag on the plants. This is important because the plants with this plastic
response are taller and therefore extend higher in the water column to increase light
interception. Conversely, this mechanism would not function in relatively high flow
environments because the plants would be deflected downwards (i.e., increased
compliance) and/or not be robust enough to withstand high drag forces. Consequently,
the plant-flow interaction can be linked to the productivity of freshwater macrophytes as
demonstrated by the different responses to flow discussed above. These different
responses observed throughout this study suggest that the fluid dynamic environment
influences the plasticity of plant architecture, and the evolution of leaf morphology in
general.

222
Literature Cited
Ackerman, J.D., and A. Okubo. 1993. Reduced mixing in a marine macrophyte canopy.
Functional Ecology. 7:305-309.
Anderson, S.M., and A.C. Charters. 1982. A fluid dynamics study of seawater flow
through Gelidium rudifrons. Limnology and Oceanography. 37(3):399-412.
Borchardt, M.A. 1994. Effects of flowing water on nitrogen- and phosphorus-limited
photosynthesis and optimum N:P ratios by Spirogyra fluviatilis (Charophyceae).
Journal o f Phycology. 30:418-430.
Cronin, G., and D.M. Lodge. 2003. Effects of light and nutrient availability on the
growth, allocation, carbon/nitrogen balance, phenolic chemistry, and resistance to
herbivory of two freshwater macrophytes. Oecologia. 137:32-41.
Ennos, A.R. 1999. The aerodynamics and hydrodynamics of plants. The Journal o f
Experimental Biology. 202:3281-3284.
Gutierrez, L.M., and C. Femandez. 1992. Water motion and morphology in Chondrus
crispus (Rhodothyta). Journal o f Phycology. 28:156-162.
Gutschick, V.P. 1999. Research reviews: Biotic and abiotic consequences of differences
in leaf stracture. New Phytologist. 143(1):3-18.
Henry, H.A.L., and S.C. Thomas. 2002. Interactive effects of lateral shade and wind on
stem allometry, biomass allocation, and mechanical stability in Abutilon theophrasti
(Malvaceae). American Journal o f Botany. 89(10): 1609-1615.
Hurd, C.L., P.J. Harrison, and L.D. Dmehl. 1996. Effect of seawater velocity in
inorganic nitrogen uptake by morphologcially distinct forms of macrocystis
intgrifolia from wave-sheltered and exposed sites. Marine Biology. 126:205-214.
Hurd, C.L. and C.L. Stevens. 1997. Flow visualization around single- and multiplebladed seaweeds with various morphologies. Journal o f Phycology. 33:360-367.
Hurd, C.L, C.L Stevens, B E. Laval, G.A. Lawrence, and P.J. Harrison. 1997.
Visualization of seawater flow around morphologically distinct forms of the giant
kelp macrocystis integrifolia from wave-sheltered and exposed sites. Limnology and
Oceanography. 42(1).T56-163.
Idestam-Almquist, J., and L. Kautsky. 1995. Plastic responses in morphology of
Potamogeton pectinatus L. to sediment and above-sediment conditions at two sites in
the northern Baltic proper. Aquatic Botany. 52:205-216.

223

Nepf, H.M. 1999. Drag, turbulence, and diffusion in flow through emergent vegetation.
Water Resources Research. 35(2):479-489.
Nowell, A.R.M., and P.A. Jumars. 1984. Flow environments of aquatic benthos. Annual
Review o f Ecology and Systematics. 15:303-328
Sand-Jensen, K. 2003. Drag and reconfiguration of freshwater macrophytes. Freshwater
BWogy. 48:271-283.
Santamaria, L. 2002. Why are most aquatic plants widely distributed? Dispersal, clonal
growth and small-scale heterogeneity in a stressful environment. Acta Oecologica.
23:137-154.
Schutten, J., and A.J. Davy. 2000. Predicting the hydraulic forces on submerged
macrophytes from current velocity, biomass and morphology. Oecologia. 123:445452.
Sculthorpe, C D . 1967. The Biology o f Aquatic Vascular Plants. Edward Arnold, London
(Reprinted 1985 Koeltz Scientific, Konigstein), 610 pp.
Sheath, R.G., and J.A. Hambrook. 1988. Mechanical adaptations to flow in freshwater
red algae. Journal o f Phycology. 24:107-111.
Speck, O. 2003. Field measurements of wind speed and reconfiguration in Arundo donax
(Poaceae) with estimates of drag forces. American Journal o f Botany. 90(8): 13531256.
Vogel, S. 1994. Life in Moving Fluids The Physical Biology o f Flow. Princeton
University Press, Princeton, N.J. 467 pp.
Wright, J.P., W.S.C. Gurney, and C.G. Jones. 2004. Patch dynamics in a landscape
modifies by ecosystem engineers. Oikos. 105:336-348.