Content uploaded by Nigel Willby
Author content
All content in this area was uploaded by Nigel Willby on Aug 01, 2018
Content may be subject to copyright.
THE ATRIUM, SOUTHERN GATE, CHICHESTER, WEST SUSSEX P019 8SQ
***IMMEDIATE RESPONSE REQUIRED***
Your article may be published online via Wiley's EarlyView® service (www.interscience.wiley.com) shortly after receipt of corrections.
EarlyView® is Wiley's online publication of individual articles in full-text HTML and/or pdf format before release of the compiled print issue of
the journal. Articles posted online in EarlyView® are peer-reviewed, copy-edited, author-corrected, and fully citable via the article DOI (for
further information, visit www.doi.org). EarlyView® means you benefit from the best of two worlds - fast online availability as well as
traditional, issue-based archiving. Please follow these instructions to avoid delay of publication
READ PROOFS CAREFULLY
• This will be your only chance to review these proofs. Please note that once your corrected article is posted online, it is considered legally
published, and cannot be removed from the Web site for further corrections.
• Please note that the volume and page numbers shown on the proofs are for position only.
ANSWER ALL QUERIES ON PROOFS (Queries for you to answer are attached as the last page of your proof.)
• List all corrections and send back to the editor as detailed in the covering e-mail, by fax or mail.
CHECK FIGURES AND TABLES CAREFULLY
• Check size, numbering, and orientation of figures.
• All images in the PDF are downsampled (reduced to lower resolution and file size) to facilitate Internet delivery. These images will appear at
higher resolution and sharpness in the printed article.
• Review figure legends to ensure that they are complete.
• Check all tables. Review layout, title, and footnotes.
COMPLETE CTA (if you have not already signed one)
• Please send a scanned copy and post your completed original form to the production contact for the journal - address and e-mail address
detailed in the covering e-mail. We cannot publish your paper until we receive the original signed form.
OFFPRINTS
• 25 complimentary offprints of your article will be dispatched on publication. If your address has changed, please inform the production contact
for the journal – address and e-mail address detailed in the covering e-mail.
Additional reprint and journal issue purchases
• Additional paper reprints (minimum quantity 100 copies) are available on publication to contributors. Quotations may be requested from
author_reprints@wiley.co.uk. Orders for additional paper reprints may be placed in advance in order to ensure that they are fulfilled in a timely
manner on publication of the article in question. Please note that offprints and reprints will be dispatched under separate cover.
• PDF files of individual articles may be purchased for personal use for $25 via Wiley’s Pay-Per-View service (see
http://www3.interscience.wiley.com/aboutus/ppv-articleselect.html).
• Please note that regardless of the form in which they are acquired, reprints should not be resold, nor further disseminated in electronic or print
form, nor deployed in part or in whole in any marketing, promotional or educational contexts without further discussion with Wiley. Permissions
requests should be directed to permreq@wiley.co.uk
• Lead authors are cordially invited to remind their co-authors that the reprint opportunities detailed above are also available to them.
• If you wish to purchase print copies of the issue in which your article appears, please contact our Journals Fulfilment Department cs-
journals@wiley.co.uk when you receive your complimentary offprints or when your article is published online in an issue. Please quote the
Volume/Issue in which your article appears.
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
RIVER RESEARCH AND APPLICATIONS
River Res. Applic. 21: 1–18 (2005)
Published online in Wiley InterScience
(www.interscience.wiley.com). DOI: 10.1002/rra.917
CHANNEL DYNAMICS AND GEOMORPHIC VARIABILITY AS CONTROLS
ON GRAVEL BAR VEGETATION; RIVER TUMMEL, SCOTLAND
DAVID GILVEAR* and NIGEL WILLBY
School of Biological and Environmental Sciences, University of Stirling, FK9 4LA, Scotland
ABSTRACT
This paper presents the results of an investigation into environmental controls on vegetation dynamics on gravel bars. Such
environments are a hotspot of threatened plant biodiversity and the dynamics of their vegetation reflect a range of processes
that should be indicative of the integrity of the wider floodplain ecosystem. The study was undertaken on a 2 ha mid-channel
gravel bar complex that evolved over two decades, in response to several high magnitude flood events (including two with a
return period in excess of 25 years), on a ‘wandering’ reach of the River Tummel, Scotland. Over 180 plant species, including a
number of national or regional scarcities, had colonized. The fluvial chronology of the site was documented via sequential sets
of aerial photography that revealed a number of discrete surfaces created by individual floods. Environmental heterogeneity,
both within and between fluvial units, was investigated by field sampling of vegetation and abiotic variables at 66 locations. The
fluvial surfaces were assigned to five habitat classes that ranged in age from two to approximately 20 years, from fine gravel to
cobbles, and maintained an elevation range of up to 2.5 metres above low flow river levels. Multivariate analysis highlighted the
relative importance of elevation, grain size, moisture content and infiltration and trapping of fines in controlling plant species
composition. After standardizing sampling effort the habitat mosaic was found to support on average 1.36 times more species
than an equivalent sample of any one habitat. In terms of biodiversity and river management, our results emphasize the impor-
tance of sustaining fluvial processes that preserve the habitat mosaic in order to conserve the characteristic biota of gravel bar
complexes and river channel islands. Copyright #2005 John Wiley & Sons, Ltd.
key words: gravel bars; riparian vegetation; plant diversity; habitat mosaic; gravel bed rivers; channel change; islands
INTRODUCTION
Fluvially-derived surfaces are the template for the establishment of riparian plant communities. An understanding
of the interaction between such surfaces and their vegetation is not only of scientific interest but also essential to
the successful reinstatement of more natural riparian communities on heavily modified river corridors and the
restoration of associated biodiversity and function. Bars are an integral landform of many rivers (Poff and Ward,
1990; Edwards et al., 1999; Gurnell and Petts, 2002). The extent of vegetation cover on bars is related to the
amount of time the bar surface has been exposed above the seasonal low-water mark (since this regulates moisture
and fertility stress), the distance to the water table, the physical nature of the sediments and their stability and the
biological traits of potential plant colonists. Depending on these factors, newly formed bars are progressively vege-
tated as they accrete vertically and laterally and it thus becomes difficult to define where a point bar becomes part
of the floodplain and a mid-channel bar becomes an island; islands are considered to be well vegetated in this study.
The separation is in effect artificial but the fact that bar development is often the first stage in island creation and
floodplain evolution is important both geomorphologically and ecologically.
In having an ecotonal position between aquatic and terrestrial environments bars have the potential to support an
unusually varied biota and range of environmental processes (Naiman and Dechamps, 1997). In the UK, for exam-
ple, the importance of exposed riverine sediments for ground beetle assemblages is now recognised (Eyre et al.,
Received 24 August 2004
Revised 21 July 2005
Copyright #2005 John Wiley & Sons, Ltd. Accepted 15 August 2005
*Correspondence to: Dr David Gilvear, School of Biological and Environmental Sciences, University of Stirling, FK9 4LA, Scotland, UK.
E-mail: d.j.gilvear@stir.ac.uk
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
2001; Sadler et al., 2004). A recent analysis of patterns of rarity in British plants (Pilgrim et al., 2004) has also
confirmed that threatened species are over represented on river gravels, spits and lakeshores, as well as arable and
horticultural, montane, inland rock and supra-littoral sediment habitats from which the flora of gravel bars is often
drawn. One interpretation of their ecotonal position is that the vegetation dynamics on islands and bars is effec-
tively an indicator of more general floodplain ecosystem health (Ward and Tockner, 2001). However, due to a com-
bination of flood control engineering, impoundments, gravel mining and navigation, islands represent an
increasingly endangered attribute of floodplain corridors (Gurnell and Petts, 2002). Consequently there is an urgent
need to assess the functional significance of these landforms at a floodplain level.
Simons and Simons (1987) view river bars as large bedforms resulting from deposition. Their classification and
evolution has been the subject of many investigations by geomorphologists. For example, on meandering rivers,
depositional features on the inside of bends and attached to the floodplain have been identified as point bars. Their
evolution has been mapped by using the pattern, often identified via the distribution of vegetation, of floodplain
ridge and swale left by migrating point bars (Nanson and Cloke, 1992). Mid-channel bars are more typical of
braided and wandering planforms and tend to be unvegetated due to rapid evolution and frequent bed instability.
Brice (1964) defined mid channel bars as being unvegetated and submerged at bankfull whereas islands are vege-
tated and emergent at bankfull stage. In reality this distinction is blurred by the fact that during the low water
season annual plant species can establish on otherwise geomorphologically active gravel bars while mid-channel
bars are often a response to gravel accumulation in the lee of floodborne trees stranded during falling stage (Gurnell
et al., 2001). In addition to point bars and mid-channel bars numerous other types have been identified primarily
relating to their in-channel location but also sediment composition. Thus a major distinction is made between sand
and gravel bars.
A simple model of bar development was proposed by Jaeggi (1987). In the first phase sediment is deposited until
a limiting height is achieved. In phase two, material is deposited in the lee of the initial phase as a tail. However,
bars should not be thought of as single morphological/sedimentological entities. Most exist for time periods in
excess of a single flood and their morphology is therefore often the result of a complex history of erosional and
depositional modification linked to the nature of the flood series following bar initiation. Discrete morphological
units such as bar heads, bar lobes, avalanche faces, bar tails, cross-bar channels and sloughs can thus be identified.
Consequently many bars show considerable internal topographic, sedimentological and chronological variability.
Despite numerous botanical studies of floodplain and riparian landscapes (e.g. Barnes 1978; Menges and Waller,
1983; Kalliola and Puhakka, 1988; Prach, 1994; Tabbacchi et al., 1996; Bornette et al., 1998a; Girel and Manne-
ville, 1998; Piegay et al., 2000; Gilvear et al., 2000; Gurnell et al., 2001) there has been little explicit focus on bars
or exposed sediments, particularly on gravel bed rivers. Recent work in the USA has focussed on the distribution of
pioneer tree seedlings on sand bars (Robertson and Augspurger, 1999; Dixon et al., 2002). For example, Bendix
and Hupp (2000) working on the gravel bars of the Missouri River established that there was a critical linkage
between floods, propagule transport, bar formation and cottonwood forest development and riparian species rich-
ness. On Passage Creek, USA, Hupp (1983) identified four vegetation zones on bars and islands that reflected the
frequency of flooding and flood damage. In protected areas of the channel, patches of Salix nigra,Orontium aqua-
tica and Justica americana rooted below the water line. Close to the channel there was an herbaceous zone with no
woody species. Further from the waters edge a shrub zone was present (e.g. Alnus serrulata); here the plants
sprouted from woody debris. In between the latter two habitats was a zone that contained species from both. Simi-
larly, Dykaar and Wigington (2000) working on the Willamette River in Oregon traced patterns of cottonwood
floodplain forest back to the evolution of underlying bar forms. In the case of the 22 km reach of the Willamette
River under investigation an 80% decline in bar and island area between 1910 and 1988 due to river management
activities was also observed. In light of this, the researchers concluded that restoration of the potential of the river
to form gravel bars was critical to restoring the river-floodplain system. In Europe, the interaction between vegeta-
tion and bar and island formation has been investigated but to date this has been restricted principally to recent
work on the alpine Tagliamento river with a specific focus on propagule forms and establishment of woody species
(e.g. Kollman et al., 1999; Karrenberg et al., 2003; Francis et al., In press).
Instream and riparian plant species diversity depends on the one hand on the influence of scale-dependent pro-
cesses including disturbance (i.e. flooding, channel planform instability, erosion and sedimentation; Cui et al.,
2000; Vervuren et al., 2003), physical stresses (i.e. drought or waterlogging; Ernst, 1990; Capon, 2003) and biotic
2D. GIL VEAR AND N. WILLBY
Copyright #2005 John Wiley & Sons, Ltd. River Res. Applic. 21: 1–18 (2005)
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
interactions (i.e. interspecific competition for light or nutrients; Menges and Waller, 1983; Willby et al., 2001 or
herbivory; Elger et al., 2004), and on the other hand on population dynamics, and patterns of seed dispersal and
vegetative regeneration (Marston et al., 1995; Hughes, 1997; Abernethy and Willby, 1999). Salo et al. (1986) for
example considered the fluvial dynamics of river systems to be fundamental in creating and maintaining high plant
diversity. The resulting landforms influence patterns of vegetation succession through a multitude of controls
including elevation and hence frequency, depth and duration of flooding, distance to the water table and various
soil properties, in particular particle size and moisture content (Hupp and Osterkamp, 1985; Prach, 1994; Wondzell
et al., 1996; Robertson and Augspurger, 1999; Lyon and Sagers, 2003). These studies and earlier work such as that
of Barnes (1978) on the floodplain forests of the River Wisconsin tend to find elevation to be the best explanatory
variable for herbaceous species distribution. Moreover, by creating fluvial surfaces of differing age but close phy-
sical similarity discrete plant communities exist in close juxtaposition reflecting different seral stages. In essence it
is likely that physical habitat heterogeneity and fluvial disturbance are the over-riding local controls on gravel-bar
plant diversity. Their role however, is conditional upon the composition of the regional species pool in terms of
plant traits and the attendant effects of population fragmentation and site connectivity on opportunities for disper-
sal and colonisation from lateral and upstream sources. The importance of understanding the relative roles of the
various abiotic controls on floodplain vegetation and therefore being able to model the contribution of channel
change to floodplain vegetation dynamics, has recently been highlighted (Ward and Tockner, 2001; Richards
et al., 2003). Coupled geomorphological-ecological models sensitive to the effects of evolving channel morpho-
logy and levels of fluvially induced instability, are needed for river restoration and management purposes.
The results presented here are part of a programme of research on riverine landscape diversity on the River
Tummel, Scotland. The project was set-up to investigate the role of fluvial-derived heterogeneity on habitat and
plant diversity adjacent to wandering gravel bed rivers and the findings of linked research have been reported
elsewhere (Gilvear and Winterbottom, 1998; Parsons and Gilvear, 2002; Gilvear et al., Submitted). The primary
purpose of this paper is to identify the major controls on vegetation colonisation and plant diversity on gravel bars
and their implications for riparian plant diversity.
THE STUDY AREA
The River Tummel downstream of the small town of Pitlochry, Scotland, is a sizeable river by UK standards
(Figure 1A). Typically it is 60 meters wide and flows within a wandering gravel-bed channel (e.g. Gilvear and
Winterbottom, 1998; Bryant and Gilvear, 1999; Winterbottom, 2000) for a distance of 10 km until its confluence
with the River Tay. Here the River Tummel has a mean discharge of 70 m
3
s
1
. The highest peak flow recorded
since gauging started in 1952 was 1048 m
3
s
1
in March 1993. The next three largest recorded floods occurred in
March 1989, February 1990 and March 1994. This high frequency of large flood events in the past decade relates to
non-stationarity of climate reflected in an increase in winter rainfall over the last two decades. Increased flooding
since 1988 has been noted elsewhere in Scotland (Black, 1996; Black and Hardie, 2001). The flows come from a
catchment area of 1649 km
2
when measured at Pitlochry and relate to an average annual precipitation total of over
1700 mm in the headwaters. Low evapotranspiration, steep slopes, thin soils and impermeable catchment geologies
also contribute to the high runoff per unit area. The runoff regime is naturally flashy and can include a significant
snow-melt contribution given the northerly latitude and location of much of the drainage basin within the southern
Grampian mountains. The geology of the catchment consists of resistant metamorphic rocks of the Upper
Dalradian epoch. Within the Piedmont valley floors, outside of bedrock reaches, extensive drifts of fluvial and
lacustrine alluvium cover the glacially scalloped floors. The valley fill deposits are highly variable in nature
and therefore have variable resistance to fluvial erosion. In these areas, small tributary alluvial fans can also control
lateral river channel movement.
The reach in which this study took place lies some 8 kilometres downstream of the Faskally Dam at Pitlochry
(Figure 1B), at an elevation of 60 m and opposite an area known as Ballinluig Island. In this reach a gravel bar
complex has evolved over the last 25 years with discrete geomorphic units, each in part being attributable to the
geomorphic impact of a series of large floods that has occurred in recent years. A large percentage of the now
partially colonized surfaces were bare gravel in 1993. One area in 1990, and for a number of years preceding this
date, was well vegetated but most of the island was over-ridden by gravels during floods in the late 1980s and early
Q1
CONTROLS ON GRAVEL BAR VEGETATION
Q1
3
Copyright #2005 John Wiley & Sons, Ltd. River Res. Applic. 21: 1–18 (2005)
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
1990s. The entire gravel bar complex is less than 2 hectares in size. The nature conservation value of the river and
adjacent floodplain in this reach was recognised in 1981 by the Nature Conservancy Council for Scotland (now
Scottish Natural Heritage) who designated the area as a Site of Special Scientific Interest (SSSI). The river is also
protected under the European Habitats Directive through designation as a Special Area for Conservation (SAC),
and is managed as a wildlife reserve by the Scottish Wildlife Trust. These designations are based on the presence of
a series of extensive gravel bars in various stages of colonisation from bare gravel to mixed woodland and includ-
ing abandoned channels and floodplain forest. The gravel bars are breeding sites for birds characteristic of open
wetland habitats including Ringed Plover (Charadrius hiaticula) and Common Tern (Sterna hirundo).
METHODOLOGY
Field studies
Twelve transects were set up across the gravel bar complex to encompass the lateral and longitudinal, morpho-
logical, sedimentological and floristic variability. Sampling points were spaced approximately 8 metres apart along
each transect; in total 66 sampling points were set-up. All sample points were surveyed with x,y,z co-ordinates
Figure 1. (A) The River Tay and Tummel catchment and (B) location of study reach
Q7
Q7
4D. GIL VEAR AND N. WILLBY
Copyright #2005 John Wiley & Sons, Ltd. River Res. Applic. 21: 1–18 (2005)
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
recorded. Elevation was recorded both as relative to the waters edge at the furthest upstream point on the exposed
bar surface and also relative to the water’s edge at the start of each transect; allowing correction for water slope
(average slope over reach was 1:120).
The vegetation was surveyed and recorded in the field within a 22 m quadrat at each of the 66 sampling points;
this size was chosen as it had previously been found to be optimum for gravel bar habitats on the river (Parsons and
Gilvear, 2002). All species present were recorded and their abundance assessed as percentage cover. All sampling
was undertaken during two dry days in late July 2002. The grain size of the surface layer of the exposed bar sedi-
ments was quantified by randomly sampling 100 particles within the 2 m 2 m quadrat, particles being measured
using a graduated ‘pebble plate’. The surface layer was also cleared over a 0.3 0.3 m area and a 2–4 kg sample of
the underlying substrate collected to a maximum depth of 10 centimetres. The sample was initially passed through
a 32 mm sieve in the field to remove the coarsest sediment fraction and then sealed for laboratory analysis.
The gravel bar complex was first visited immediately following the large flood event that occurred in January
1990 and thereafter was visited annually. A detailed study on the bar morphology and pattern of grain size varia-
bility was also undertaken in 1994. This data and the annual visits were used to interpret the evolution of the bar
complex from available aerial photography.
Laboratory analysis
Samples of the sub-surface gravels were analysed for grain size distribution, moisture and organic content. Soil
moisture content by percentage weight was quantified by drying at 110C for two hours. Organic matter by per-
centage weight was determined from loss on ignition at 550C. Particle size analysis was undertaken by dry sieving
down to 63 mm (clays in the gravels within this reach are absent).
Aerial photographic interpretation
The morphology and evolution of the gravel bar complex was described primarily by analysis of sequential sets
of aerial photography for 8th May 1946, 7th August 1968, 6th July 1971, 15th May 1988, 10th July 1992, 31st
January 1993, 13th June 1994 and 28th July 1999. Photographic scales varied between 1:10 000 and 1:2500 with
individual dates. The aerial photographs were referenced to each other using the GIS system ArcInfo. This process
involves the warping of an image to a set of ten reference points identifiable on all of the aerial photographs. An
error margin of approximately þ/2 metres was determined. The bar morphology was also surveyed in 1994
(Cameron, 1996; Figure 2).
Data analysis
The environmental data was classified into five groups using cluster analysis (based on average linkage) within
MINITAB v 12. Elevation, mean particle size and sorting of the armoured layer, percentage soil moisture content,
percentage organic matter, and sediment particle size formed the input variables. The distribution of the five result-
ing clusters over the site is illustrated in Figure 3. The mean sample species richness, and the frequency and abun-
dance of plant species in each of the five clusters were then calculated. Species unique to each cluster were also
identified. To estimate the species pool associated with each cluster based on a constant sampling effort, extrapo-
lation and rarefaction, as appropriate to the number of samples per cluster, were undertaken using the programme
Estimate S v 5.01 (Colwell, 1997). This enabled a comparison of species turnover (beta diversity) between samples
within each of the different clusters and an assessment of the relative contribution of each cluster to the species
pool associated with the mosaic of habitats on the site as a whole.
Species-environment relationships were analysed using Canonical Correspondence Analysis (CCA) with
CANOCO v 4.5. The influence of variables was tested using manual forward selection supported by Monte Carlo
random permutation tests (999 runs) of the significance of each additional variable.
RESULTS
Morphology and evolution of the gravel bar
In the study reach between 1945 and 1971 a mid-channel partially vegetated gravel bar attached to the right river
bank was apparent under summer low flows (Figure 4A). Upstream, over a distance of 300 metres, two separate
CONTROLS ON GRAVEL BAR VEGETATION
Q1
5
Copyright #2005 John Wiley & Sons, Ltd. River Res. Applic. 21: 1–18 (2005)
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
unvegetated mid-channel bars are apparent on the 1971 photography. In 1988 channel switching occurred in the
study reach with a bar attached to the left bank at its upstream end. The bar morphology essentially consisted of a
bar head and single bar tail, the latter being partially vegetated and forming a small backwater between it and the
left bank.
By 1992 the morphology of the bar had changed significantly (Figure 4B). This change was the result of bedload
deposition during the large flood event that occurred in 1990. Upstream the earlier morphology had been replaced
by a higher and larger gravel lobe with a marked avalanche face of two metres elevation at the downstream end of
the landform. In the lee of the gravel lobe a second bar tail form was also created giving the bar complex a two-
tailed morphology. The pre-1992 bar surface not overridden during the 1990 event by the gravel lobe was vegetated
Figure 2. The morphology of the gravel bar complex (after Cameron, 1996); Elevations are recorded as height above the water surface at the
downstream end of the bar complex
6D. GIL VEAR AND N. WILLBY
Copyright #2005 John Wiley & Sons, Ltd. River Res. Applic. 21: 1–18 (2005)
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
with grasses and trees. Where the gravel lobe engulfed the highest elevations of the earlier bar form some vegeta-
tion protruded through the new gravel surface.
The 1993 flood event caused further modification to the bar morphology as evidenced by field observation and
the 1994 aerial photography (Figure 4C). The gravel lobe avalanche face advanced by up to ten metres and
accreted laterally towards the right bank. A third bar tail also formed as a result of the further lateral accretion
to the bar lobe. Since 1994 the bar has maintained essentially the same geomorphology although a small amount
of lateral accretion has widened the gravel lobe and added a fourth discrete bar tail. Annual flooding and bedload
movement however, has merged the surface of the third and fourth bar surface into a single geomorphic unit.
Presently the bar is 400 m long and 85 m wide at its maximum. Its maximum elevation range is 2.8 m. The bar
head ranges from between a few centimetres and 2.5 m above summer low flow. The bar tails lie at an elevation of
between 1.5 and 2.0 m below the maximum elevation of the bar head and up to 0.50 to 0.65 m above summer low
Figure 3. The spatial distribution of the 5 habitat groups obtained by cluster analysis. Base photograph as in Figure 4D. Numbers refer to
vegetation cluster as in Table 1, Appendix 1 and the text
CONTROLS ON GRAVEL BAR VEGETATION
Q1
7
Copyright #2005 John Wiley & Sons, Ltd. River Res. Applic. 21: 1–18 (2005)
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
flow. (Table I). Mean sediment size for the armour layer of gravel samples varied between 21 and 64 mm (Table I).
Eighty fourth percentile values varied between 15 and 118mm. Some gravel areas were totally draped with sand
and recorded as such. Moisture levels in the sediments ranged from less than 5% to 35%. Organic content varied
between less than 1% and 15% but values were generally below 3%.
Vegetation in relation to the habitat mosaic
The environmental characteristics of each of the 5 groups of sites identified by the cluster analysis are shown in
Table I. Cluster 1 was found in mid- elevation areas with dry, coarse, poorly-sorted sediments and supported sparse
pioneer vegetation. Cluster 2 sites were found at the highest elevation and like cluster 1 had coarse, poorly-sorted
sediments with a slightly higher moisture content due to higher organic content. The area supported a diverse vege-
tation community containing a mix of herbs, bryophytes (particularly Racomitrium), and grasses with broom
(Sarothamnus scoparius) and gorse (Ulex europaeus) scrub and occasional sycamore saplings (Acer pseudoplata-
nus). Cluster 3 sites were at mid elevations, very dry and of moderate grain size. They supported mixed, herb-rich
mesotrophic grassland. Cluster 4 comprised finer, well-sorted, damp gravels that made up the bar tail areas. They
Figure 4. The geomorphic evolution of the gravel bar complex (A) 1946; (B) 1989; (C) 1994; (D) 1999. Flow is from right to left and the reach
500 metres in length
8D. GIL VEAR AND N. WILLBY
Copyright #2005 John Wiley & Sons, Ltd. River Res. Applic. 21: 1–18 (2005)
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
supported mixed herb/shaded neutral grassland. Cluster 5 contained wet organic low-lying sediments and were
found primarily on the lee side of the individual bar tails and where sand splays were developed in the lee of willow
(Salix spp.) bushes. Luxuriant vegetation was prevalent with a composition typical of the margins of palaeochan-
nels in the early stages of abandonment. The distribution of these five clusters over the site is illustrated in Figure 4.
The overall bar vegetation was dominated by the following species: Agrostis canina, Arrhenatherum elatius,
Bryum pseudotriquetrum, Centaurea nigra, Dactylis glomerata, Festuca ovina, Galeopsis tetrahit, Holcus lanatus,
Impatiens glandulifera, Leontodon autumnalis, Ononis repens, Plantago lanceolata, Rumex acetosella, Taraxa-
cum agg, Viola riviniana, Silene maritimus, Sarothamnus scoparius, and Senecio viscosa (Appendix I). These
twenty species accounted for over 70% of the total plant cover.
In total 181 plants species were found to be present of which 87% were noted within the 66 standard sampling
units. Plant species richness in samples was relatively high in all clusters with values ranging by almost two fold
from 10 in cluster 1 to 19 in cluster 2. Sampling effort was roughly proportional to the area covered by the indi-
vidual clusters. Sampling in the small wet areas in the lee of the bar tails thus only yielded 4 samples whereas 23
covered the central high ‘plateau’ area of the bar. To enable comparison between clusters rarefaction and in one
case (cluster 4) extrapolation was therefore used to establish the species pool that would have been associated with
each cluster given a constant sampling effort. Table I reveals that the species pool increases from cluster 1 to 4 and
is lower in 5. Turnover between samples (i.e. the ratio of species pool to average sample richness) was high in all
clusters, although somewhat lower in 2 and 5. The number of unique species was more than twice as high in cluster
2 than the average of the other 4 clusters. In total 52 species were estimated to be unique to one of the clusters.
Overall, the mosaic (with clusters in the ratio observed) supported a species pool of, on average, 1.36 times higher
than that of an equal sized sample of any one cluster. Although cluster 4 contained almost 92% of the species pool
of the whole gravel bar mosaic it had a low uniqueness value compared to cluster 2, with which most of the region-
ally rare species were associated (Figure 5). This emphasizes the importance of sustaining fluvial processes that
preserve the habitat mosaic if the conservation value of similar bar complexes is to be maintained.
Species-environment relationships
Topography and sediment characteristics reflect the outcomes of a common set of geomorphological processes
and are naturally strongly inter-correlated. The analysis of species-environment relationships demonstrated that
elevation was the most important variable. Moisture content and surface grain size, as well as distance down
the length of the bar were all significant in explaining the spatial distribution of the vegetation independent of
Table I. Environmental and botanical characteristics of each cluster
Cluster Elevation Sediment Sediment Moisture Organic Armouring Armouring
(metres) size (mm) sorting content (%) content (%) size (mm) sorting
1 0.36 1.6 1.4 2.1 1.1 55.4 27.4
2 0.73 1.7 1.4 4.1 3.1 51.4 25.9
3 0.31 1.2 1.2 3.2 1.3 27.8 12.6
4 0.32 0.3 0.1 14.5 2.0 11.5 5.2
5 0.22 0.6 0.7 34.6 10.0 0.5 0.7
Cluster Vegetation Sample No. samples Standardised Error Beta index Unique
cover (%) richness species pool species
1 20.3 10.4 11 55.6 1.09 0.53 2.9
2 45.5 19.1 16 76.5 2.14 0.40 19.5
3 30.0 14.7 23 80.7 3.50 0.55 11.8
4 49.6 18.5 12 94.7 1.39 0.51 9.8
5 50.0 18.0 4 74.0 N/A 0.40 7.0
All values are means based on the number of samples per cluster. Sorting coefficients are calculated based on (D
84
D
16
)/2 where D
x
is the
diameter of the xth percentile particle. Sample richness refers to alpha diversity (i.e. the mean number of species per 2 2 m quadrat). The
species pool for each cluster is based on a standard sample size of 10 calculated using rarefaction (Coleman estimate) or extrapolation (cluster 5
only). The beta index, a measure of spatial turnover between samples, ¼species pool/(sample richness 10). The number of unique species is
based on the number of species observed only in one cluster scaled relative to the size of the estimated species pool.
CONTROLS ON GRAVEL BAR VEGETATION
Q1
9
Copyright #2005 John Wiley & Sons, Ltd. River Res. Applic. 21: 1–18 (2005)
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
elevation alone (Table II). Together these represent the most parsimonious set of variables and would be a suitable
subset of predictors for subsequent studies at other sites. Both elevation and distance down the bar are strictly
spatial variables that provide a surrogate for one or more deterministic factors (e.g. duration of inundation) and
may potentially convey much of the variation in other measured parameters such as soil moisture. However, when
the influence of elevation and distance down the bar was tested after first fitting all the substrate related parameters
it remained highly significant in both cases ( p0.001). The latter variable together with the control of elevation
most likely reflects selective ‘entrapment’ of fine floodborne sediment and propagules.
DISCUSSION
Geomorphological controls on vegetation composition
In common with many previous studies of river floodplains elevation proved to be the key explanatory variable
for vegetation composition (e.g. Shelford, 1954; Barnes, 1978; Menges and Waller, 1983; Hupp and Osterkamp,
Figure 5. Standardised species richness, uniqueness and percentage of the total species pool for the 5 clusters and overall gravel bar mosaic
Q7
Table II. Species-environmental relationships analysed using CCA
Variable Marginal Effects Conditional Effects
TVE P Input order FVE F P
Elevation 5.9 0.001 1 23.7 3.99 0.001
Armour sort 4.8 0.001 5 8.9 1.60 0.038
Armour size 4.6 0.001 4 12.6 2.36 0.001
% moisture 4.6 0.001 2 13.3 2.36 0.001
Long axis 4.6 0.001 3 13.2 2.28 0.001
Sediment size 4.4 0.001 7 7.2 0.95 0.520
Sediment sort 3.7 0.001 6 7.4 1.24 0.140
Lateral axis 2.6 0.010 ns
% organic 2.2 0.038 ns
All variables 21.0 0.001
Marginal effects assess the influence of each variable, including its covariation with all other variables. The conditional effects consider vari-
ables as supplied by a forward selection procedure that ranks variables in accordance with their ability to explain initial and residual variation
(see input order). Those variables significant at p0.05 based on Monte Carlo random permutation tests (n¼999) and Bonferroni correction
are shown in bold. This can be regarded as a minimum adequate model for representing variation in species composition at the site. TVE ¼%
Total Variance Explained; FVE ¼Fraction Variance Explained by all independent variables.
Q7
10 D. GIL VEAR AND N. WILLBY
Copyright #2005 John Wiley & Sons, Ltd. River Res. Applic. 21: 1–18 (2005)
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
1985; Lyon and Sagers, 1998). Elevation affects the frequency and duration of inundation and may influence the
likelihood of bed load movement, the latter depending additionally on particle size and shape. In this study eleva-
tion over-rode the importance of age structure. For example, the three bar tails were formed at separate dates
between 1989 and 1997 but their plant species composition was very similar. This relates to the fact that they
all experience frequent bed disturbance due to their low elevation and movement of medium and fine gravel sizes
must take place particularly in the cross-bar chutes. This effectively removes the variation in age between the dif-
ferent surfaces even though each bar tail formed during a discrete flood event. However, while at lower elevations
age of bar does not relate to vegetation community type due to frequent disturbance, at higher elevations and thus
in more stable conditions, age once again becomes an important variable influencing community structure and
species composition. Plant colonisation of low-lying surfaces is probably facilitated by small floods spilling across
them and conveying vegetative propagules and seeds that are either trapped in the gravels or settle out in their
lee. The significance of hydrochory for seed dispersal has long been recognised (Anderson and Nilsson, 1999;
Jansson
Q2
et al., 2000a; Goodson et al., 2002). This may explain why distance across and downstream along
the bar were significant explanatory variables. The three bar tails were also critical in that they created slack water
areas in their lee that allowed the development of vegetation associated with cluster 5. Deposition of fine sediment
in these slack water areas was also accentuated by the establishment of dense Salix scrub.
The vegetation of the five habitat clusters represents the transition from perennial water to mature floodplain
habitat. We can therefore propose a conceptual model of the environmental controls on plant diversity and succes-
sion on gravel bars. The morphology and the sedimentology of the bar creates spatial heterogeneity and a mosaic of
habitats. In this context the most important variable is elevation, although particle size is also important. The two
together control the moisture regime of a patch; sheltered low-lying areas with fine sediments are wetter (and prob-
ably more fertile) while coarse sediments at higher elevations have a lower moisture retaining capacity. Plants in
low-lying areas are subject to low drought stress but high disturbance due to frequent inundation, deposition of
fines, and bedload movement; the later processes opening up areas for pioneer species to establish. Periodic dis-
turbance of these areas also prevents competitive displacement of pioneer species. Higher up the bar on freely
draining gravels, particularly where particle size is large and there is little interstitial sediment, only relatively
drought-tolerant species more typical of coastal shingle or montane scree environments persist. These species
are typically shade-intolerant, slow growing, with very deep roots, and prostrate evergreen foliage, but since
drought and low fertility restrict the vigour of tall perennial species the upper plateau also supports various small
summer annuals with a high output of wind-dispersed seeds. Thus xerophytic species can survive in riverine sys-
tems in such specific locations. Dessication leading to mortality of seedlings in such environments have been
reported (e.g. Sacchi and Price, 1992; Rood et al., 1998; Johnson, 2000). In these areas gradual establishment
of deep-rooting trees or nitrogen-fixing shrubs provides shade and litter and, together with settlement of large
woody debris, traps fine sediment thereby aiding moisture retention and soil development that is critical in facil-
itating establishment of other species. The succession to larger grasses and herbs, shrubs and canopy forming trees
marks a common shift from nutrient and drought limitation to light limitation across successional gradients. The
associated transition in functional attributes of plants is a common theme of studies of riparian vegetation else-
where (e.g. Hupp, 1983; Menges and Waller, 1983; Lyon and Sagers, 2003).
Within the conceptual model described above, Ballinluig Island’s particularly high species richness can be
attributed partly to high elevational variability at a coarse scale, with extensive surfaces at high, medium and
low levels and hence an overall average elevation above normal water level (Figure 6A). Bar forms with solely
high, medium or low surface elevations would obviously not yield such diversity regardless of the age of these
surfaces. The high level of species uniqueness at each elevation and high turnover between patches reflects finer
scale heterogeneity in resources and the tendency for drought stress at upper elevations or physical disturbance at
lower elevations to prevent competitive exclusion. Lyon and Sagers (1998) and Dixon et al. (2002) also discuss the
hierarchical effects of scale on riparian plant assemblages.
Figure 6B provides a schematic diagram of the seral sequences of the five plant habitat clusters identified in
relation to drought stress, particle size, topography and the aggradation of the various fluvial surfaces with time.
The timescales suggested, in the range 1 to 20 years, are based on our knowledge of the chronology of the study
site, while those beyond 20 years are estimates based on observations within other floodplain sectors along
the river. This scheme is reliant on two basic assumptions: firstly elevation increases over time (in the absence
Q2
CONTROLS ON GRAVEL BAR VEGETATION
Q1
11
Copyright #2005 John Wiley & Sons, Ltd. River Res. Applic. 21: 1–18 (2005)
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Figure 6. (A) A conceptual model of the position of Ballinluig Island in relation to European temperate zone gravel bar and floodplain island
plant diversity within a framework of elevation and elevational heterogeneity. (B) Schematic diagram of main sequences of vegetation succes-
sion on Ballinluig Island and elsewhere within the studied floodplain segment in relation to bar elevation and particle size. Numbers represent
habitat clusters identified in the present study (Table I). Boxes refer to natural climax vegetation or a mesotrophic grassland community main-
tained via grazing. Labelled arrows show the major trajectories of vegetation development with key processes and indicative time scales. Note
that at any stage within 3–5 large floods can reinitiate cluster 1 by avalanching gravels over existing communities thus creating new surfaces.
Extensive sand drapes associated with medium sized floods render clusters 3–5 especially vulnerable to invasion by non native Acer pseudo-
platanus and Impatiens glandulifera
Q7
Q7
12 D. GIL VEAR AND N. WILLBY
Copyright #2005 John Wiley & Sons, Ltd. River Res. Applic. 21: 1–18 (2005)
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
of lateral erosion) and secondly, particle size will ultimately decrease with time. Obviously during rare high mag-
nitude floods events wholesale erosion of bar segments or mobilisation of unvegetated and vegetated surface sedi-
ments is a possibility and therefore may reset the system. An incipient low level bar surface newly formed close to
the normal water level will develop a vegetation composition similar to Cluster 3. At the incipient bar margins
especially by slack water areas the community will move towards a composition similar to cluster 5. Observations
of palaeo-channels elsewhere in the floodplain (e.g. Parsons and Gilvear, 2002) indicate that the climax community
for this trajectory is wet alder woodland. With time, rapid bar aggradation, as occurred on the study site in the
flood event of 1990, leads to a new high elevation coarse surface that develops a Cluster 1 plant composition.
For Cluster 1 to progress to Cluster 2, bed stability and absence of frequent flooding is critical; this favours the
establishment of slow growing mosses, such as Racomitrium, and mat forming herbs such as Thymus, some reten-
tion of moisture and fines and the establishment of small leguminous herbs (e.g. Ononis repens) and specialist
gravel bar species such as Fillago minima,Linaria repens,Circaea alpina, and Teesdalia nudicaulis. Progression
from Cluster 2 to the climax upper terrace forest (evident across the river from the study site) depends on sand
splays and woody debris-focussed drapes that provide regeneration niches for leguminous shrubs (gorse, broom
and lupin). Provided grazing is minimal, colonisation by pine (Pinus sylvestris) and birch (Betula pendula) can
occur easily within 20 years of bar formation. Following extreme floods that cause avalanching of coarse bed
material over existing high level surfaces succession of ground layer vegetation is completely reset yet the larger
specimens of woody species can survive this partial burial and may accelerate re-colonisation by herbs. If rapid
sedimentation by fines on to cluster 3 surfaces occurs, the increase in elevation leads to a reduction in inundation
frequency and improved rooting capacity thus favouring the establishment of laterally-spreading grasses and tall
ruderal herbs that further stabilises the surface. Our observations suggest that in the presence of light grazing by
rabbits and roe deer, tree seedling establishment is prevented thus creating a pseudo-climax of tall herb, neutral
grassland. Without this control we can reach the same woodland climax as via route 1 and 2. Because these clusters
attract the deposition of fine material they are particularly vulnerable to invasion by non-native species. Here a
community dominated by Acer pseudoplatanus, an exotic woody species, and/or a riparian tall herb, Impatiens
glandulifera, can develop from clusters 3, 4 and 5, although attempts are made by the nature conservation
organisation managing the river to prevent this.
Further work is required in a number of areas to clarify the physical and biological controls on bar plant diver-
sity. Firstly, work is needed to determine the relative contribution of local versus distant (upstream) seed and pro-
pagule sources. It is likely that disruption of connectivity along rivers by dams and between rivers and their
floodplains limits plant dispersal (Bornette et al., 1998b; Ward et al., 1999; Jansson et al., 2000b; Goodson
et al., 2003). Secondly, more information is needed on the changing nature of the gravel bar environment over
the longitudinal gradient of large rivers and the extent to which it reflects the condition of the wider floodplain
corridor. One of the few examples of a study reflecting these requirements is that of Gurnell et al. (2001) on
the Tagliamento River in northern Italy. Thirdly, studies are required to provide better understanding on the pro-
cesses and controls on vegetation colonisation and establishment; for example, tolerance to flooding at critical life
history stages (Vandersman et al., 1993), causes of recruitment limitation (van Eck et al., 2005), and the role of fine
interstitial sediments in facilitating seedling establishment (Piegay et al., 2000; Goodson et al., 2003).
Significance for management of riparian corridors
The results of this study suggest that gravel bars form a highly significant component of the river corridor in
terms of supporting high plant diversity. We estimate that 70% of the terrestrial higher plant species found in a
10 km reach of the River Tummel between Pitlochry and the confluence with the River Tay can be found in a gravel
bar complex with an area of less than 2 hectares. This diversity is primarily a function of high, coarse-scale, spatial
heterogeneity in the form of distinct fluvially-derived surfaces that provide contrasting habitats for plant colonisa-
tion. In addition the opposing forces of moisture stress and fluvial disturbance, coupled with fine scale heteroge-
neity, are critical in maintaining species richness and high uniqueness at different elevations. As a result the
biodiversity of the mosaic is greater than that of an equivalent sized sample of any single cluster. It is a fundamental
ecological principle that spatial heterogeneity in limiting resources assists species coexistence (e.g. Tilman and
Pacala, 1993) but the architects of such heterogeneity are often neglected.
CONTROLS ON GRAVEL BAR VEGETATION
Q1
13
Copyright #2005 John Wiley & Sons, Ltd. River Res. Applic. 21: 1–18 (2005)
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
It is important to consider how human activities can adversely disturb gravel bars and affect their nature con-
servation value. Change in flow regime, whether natural or anthropogenically driven, is likely to significantly affect
vegetation by altering the physical characteristics of geomorphic surfaces, the abundance and composition of dis-
persing propagules, and by destroying vegetation directly. Any flow regulation that reduces bedload movement at
lower elevations, would, for example, restrict the regeneration niche of small ruderal species, particularly if the
regulation of flow also facilitates the build up of organic matter and retention of nutrients that would favour
more strongly competitive species. Some high level surfaces are only created by rare high magnitude events
(e.g. 1:50 year return period) and if lost due to activities such as channelisation, gravel mining or agricultural
improvement, may prove exceptionally difficult to restore or re-create. Reduction in the size of large floods will
also prevent the creation of high elevation, coarse bedded landforms that provide a niche for xerophytic species.
Large, medium and small sized floods are all therefore critical in maintaining the suite of landforms and processes
characteristic of natural gravel bars and islands that are important in maintaining plant diversity. Working on the
River Tagliamento Gurnell et al. (2001) and Gurnell and Petts (2002) suggest that the trajectory of biomass
production on fluvial surfaces is controlled by the interaction between the timing and severity of floods and the
capacity of different propagules to produce biomass rapidly. Thus the frequency, timing and actual flood magni-
tudes will also control vegetation via effects on seed dispersal and regeneration.
In terms of restoration of river corridors, physical placement of exposed riverine sediments and re-sculpturing
existing degraded exposures are unlikely to recreate plant diversity without full restoration of the fluvial processes
that both initiate such features and import potential colonists from the surrounding landscape. Once lost it will
prove especially difficult to re-engineer floodplain features that are the product of large events that occur on aver-
age only once every 30–50 years.
Overall this study shows that geomorphological processes provide the template for gravel bar vegetation devel-
opment (Poff and Ward, 1990) and that colonisation and succession of bare gravels and their eventual incorpora-
tion in to the floodplain is critical in maintaining high floodplain biodiversity. As such gravel bars and river islands
must be regarded as an integral component of functional river corridors not expendable features that can be
exploited by mining, deprived of sediment by dams, or simply removed in the interests of hydraulic efficiency.
CONCLUSION
This study demonstrates that gravel bars have the potential to support high plant diversity where they maintain high
morphological diversity. Morphological diversity is a major factor supporting high species richness, with elevation
above normal flow levels being the primary explanation of vegetation composition. Contrasting drought stresses
and anoxia, or flood disturbances at differing elevations across a bar, coupled with fine scale heterogeneity, also
resist competitive exclusion and promote species turnover between patches.
Overall our findings are significant in suggesting that gravel bars and river islands are a key component of the
river corridor environment and should be left ‘pristine’ where possible. Where they once occurred but are now
absent due to flow regulation or channelisation, consideration should be given to restoring a near natural flow
regime and sediment dynamics, thus enabling natural re-establishment of such features in the landscape.
REFERENCES
Abernethy VJ, Willby NJ. 1999. Changes along a disturbance gradient in the density and composition of propagule banks in floodplain aquatic
habitats. Plant Ecology 140: 177–190.
Anderson E, Nilsson C. 1999. Temporal variation in the drift of plant litter and diaspores along a small boreal river. In Relationships Between
Hydrochory and Riparian Flora in Boreal Rivers, Andersson E. Doctoral Dissertation, Umea University, Sweden.
Barnes WJ. 1978. The distribution of floodplain herbs as influenced by annual flood elevations. Wisconsin Academy of Sciences, Arts and Letters
Transactions 66: 254–266.
Bendix J, Hupp CR. 2000. Hydrological and geomorphological impacts on riparian plant communities. Hydrological Processes 14: 2977–2990.
Black AR, Hardie AM. 2001. Construction and interpretation of a 60 year flow record for the River Tay at Kenmore. Scottish Geographical
Magazine 16: 93–109.
Black AR. 1996. Major flooding and increased flood frequency in Scotland since 1988. Phys Chem Earth
Q3
20: 463–468. Q3
14 D. GIL VEAR AND N. WILLBY
Copyright #2005 John Wiley & Sons, Ltd. River Res. Applic. 21: 1–18 (2005)
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Bornette G, Amoros C, Piegey H, Tachet J, Hein T. 1998a. Ecological complexity of wetlands within a riverine landscape. Biological Conserva-
tion 85: 35–45.
Bornette G, Amoros C, Lamouroux N. 1998b. Aquatic plant diversity in riverine wetlands: the role of connectivity. Freshwater Biology 39: 267–
283.
Brice JC. 1964. Channel Patterns and Terraces of the Loup Rivers in Nebraska. US Geological Survey Professional paper 422D.
Bryant RG, Gilvear DJ. 1999. Quantifying geomorphic and riparian land cover changes either side of a large flood event using airborne remote
sensing: river Tay, Scotland. Geomorphology 29: 307–321.
Cameron D. 1996. Gravel Bar Development on the River Tummel. Unpublished, University of Stirling, BSc. thesis.
Capon SJ. 2003. Plant community responses to wetting and drying in a large arid floodplain. River Research and Applications 19: 509–520.
Colwell RK. 1997. Estimate S: statistical estimation of species richness and shared species from samples. Version 5.01. University of Connecti-
cut. http://viceroy.eeb.uconn.edu/estimates.
Cui XH, Zhong Y, Chen JK. 2000. Influence of a catastrophic flood on densities and biomasses of three plant species in Poyang Lake, China.
Journal of Freshwater Ecology 15: 537–541.
Dixon MD, Turner MG, Jin CF. 2002. Riparian tree seedling distribution on Wisconsin River sandbars: controls at different spatial scales.
Ecological Monographs 72: 465–485.
Dykaar BB, Wigington PJ. 2000. Floodplain formation and cottonwood colonisation patterns on the Willamette River, Oregon, USA. Environ-
mental Management 25: 87–104.
Edwards PJ, Kollman J, Gurnell AM, Petts GE, Tockner K, Ward JV. 1999. A conceptual model of vegetation dynamics on gravel bars of a large
Alpine river. Wetlands Ecology and Management 7: 141–153.
Elger A, Bornette G, Barrat-Segretain MH, Amoros C. 2004. Disturbances as a structuring factor of plant palatability in aquatic communities.
Ecology 85: 304–311.
Ernst EPO. 1990. Ecophysiology of plants in waterlogged and flooded environments. Aquatic Botany 38: 73–90.
Eyre MD, Luff ML, Phillips DA. 2001. The ground beetles (Coleoptera:Caribbidae) of exposed riverine sediments in Scotland and northern
England. Biodiversity and Conservation 10: 403–426.
Francis RA, Gurnell AM, Petts GE, Edwards PJ. In Press. Riparian tree establishment on gravel bars: interactions between plant growth strategy
and the physical environment. International Association of Sedimentologists Special Publication.
Q4
Gilvear DJ, Cecil J, Parsons H. 2000. Channel change and vegetation diversity on a low-angle alluvial fan: river feshie, Scotland. Aquatic
Conservation, Marine and Freshwater Ecosystems 10: 53–71.
Gilvear DJ, Willby N, Parsons H. Submitted. Geomorphological controls on vegetation patterning and plant species diversity adjacent to a
wandering gravel-bed river. River Research and Applications.
Q4
Gilvear DJ, Winterbottom SJ. 1998. Changes in channel morphology, floodplain land use and flood damage on the river Tay and Tummel:
implications for floodplain management. In United Kingdom Floodplains, Bailey R, Jose P, Sherwood BR (eds.). Westbury: Otley;
93–110.
Girel J, Manneville O. 1998. Present plant species richness of plant communities in alpine stream corridors in relation to historical river manage-
ment. Biological Conservation 85: 21–33.
Goodson JM, Gurnell AM, Angold PG, Morrissey IP. 2002. Riparian seed banks along the River Dove, UK: their structure and ecological
implications. Geomorphology 47: 45–60.
Goodson JM, Gurnell AM, Angold PG, Morrissey IP. 2003. Evidence for hydrochory and the deposition of viable seeds within winter flow-
deposited sediments: the River Dove, Derbyshire, UK. River Research and Applications 19: 317–334.
Gurnell AM, Petts GE. 2002. Island-dominated landscapes of large floodplain rivers, a European perspective. Freshwater Biology 47:
581–600.
Gurnell AM, Petts GE, Hannah DM, Smith BPG, Edwards PJ, Kollman J, Ward JV, Tockner K. 2001. Riparian vegetation and island formation
along the gravel-bed Fiume Tagliamento, Italy. Earth Surface Processes and Landforms 26: 31–31.
Q5
Hughes FMR. 1997. Floodplain biogeomorphology. Progress in Physical Geography 21: 501–529.
Hupp CR. 1983. Vegetation pattern on channel features in the Passage Creek Gorge, Virginia. Castanea 48: 62–72.
Hupp CR, Osterkamp WR. 1985. Bottomland vegetation distribution along Passage Creek, Virginia, in relation to fluvial landforms. Ecology 66:
670–681.
Jaegii MNR. 1987. Interaction of bed-load transport with bars. In Sediment transport in Gravel Bed Rivers, Thorne CR, Bathurst JC, Hey RD
(eds.). John Wiley and Sons: Chichester.
Jansson R, Nilsson C, Renofalt B. 2000a. Fragmentation of riparian flora in rivers with multiple dams. Ecology 81: 899–903.
Jansson R, Nilsson C, Dynesius M, Andersson E. 2000b. Effects of river regulation on river-margin vegetation: a comparison of eight boreal
rivers. Ecological Applications 10: 203–224.
Johnson WC. 2000. Tree recruitment and survival in rivers: influence of hydrological processes. Hydrological Processes 14: 3051–3074.
Kalliola R, Puhakka M. 1988. River dynamics and vegetation mosaicism: a case study of the River Kamjohka, northernmost Finland. Journal of
Biogeography 15: 703–719.
Karrenberg S, Kollman J, Edwards PJ, Gurnell AM, Petts GE. 2003. Patterns in woody vegetation along the active zone of a near-natural Alpine
river. Basic and Applied Ecology 4: 157–166.
Kollmann J, Vieli M, Edwards PJ, Tockner K, Ward JV. 1999. Interactions between vegetation development and island formation in the Alpine
river Tagliamento. Applied Vegetation Science 2: 25–36.
Lyon J, Sagers CL. 1998. Structure of herbaceous plant assemblages in a forested riparian landscape. Plant Ecology 138: 1–16.
Lyon J, Sagers CL. 2003. Correspondence analysis of functional groups in a riparian landscape. Plant Ecology 164: 173–181.
Q4
Q4
Q5
CONTROLS ON GRAVEL BAR VEGETATION
Q1
15
Copyright #2005 John Wiley & Sons, Ltd. River Res. Applic. 21: 1–18 (2005)
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Marston RA, Girel J, Pautou G, Piegay H, Bravard JP, Arneson C. 1995. Channel metamorphosis, floodplain disturbance and vegetation devel-
opment, Ain River, France. Geomorphology 13: 121–132.
Menges ES, Waller DM. 1983. Plant strategies in relation to elevation and light in floodplain herbs. American Midland Naturalist 122: 454–473.
Naiman RJ, Decamps H. 1997. The ecology of interfaces: riparian zones. Annual Review of Ecology and Systematics 28: 621–658.
Nanson GC, Cloke JC. 1992. A genetic classification of floodplains. Geomorphology 4: 459–486.
Parsons H, Gilvear DJ. 2002. Valley floor landscape change following almost 100 years of flood embankment abandonment on a wandering
gravel-bed river. River Research and Applications 18: 461–479.
Piegay H, Bornette G, Citterio A, Herouin E, Moulin B, Statiotis C. 2000. Channel instability as a control on silting dynamics and vegetation
patterns within perifluvial aquatic zones. Hydrological Processes 14: 3011–3029.
Pilgrim ES, Crawley MJ, Dolphin K. 2004. Patterns of rarity in the native British flora. Biological Conservation 120: 161–170.
Poff NL, Ward JV. 1990. Physical habitat template of lotic systems: recovery in the context of historical pattern of spatiotemporal heterogeneity.
Environmental Management 14: 629–645.
Prach K. 1994. Vegetation succession on river gravel bars across the northwestern Himalayas, India. Arctic and Alpine Research 26: 117–125.
Richards K, Brasington J, Hughes F. 2003. Geomorphic dynamics of floodplains: ecological implications and a potential modelling strategy.
Freshwater Biology 47: 559–579.
Robertson KM, Augspurger CK. 1999. Geomorphic processes and spatial patterns of primary forest succession on the Bogue Chitto River,
U.S.A. Journal of Ecology 87: 1052–1063.
Rood SB, Kalischuck AR, Mahoney JM. 1998. Initial cottonwood seedling recriutment following the flood of the century of the Oldman River,
Alberta, Canada. Wetlands 18: 557–570.
Sacchi CF, Price PW. 1992. The relative roles of abiotic and biotic factors in seedling demography of Arroyo Willow (Salix lasiolepis). Amer-
ican Journal of Botany 79: 395–405.
Sadler JP, Bell D, Fowles A. 2004. The hydroecological controls and conservationvalue of beetles on exposed riverine sediments in England and
Wales. Biological Conservation 118: 41–65.
Salo J, Kalliola R, Hakkinen I, Makinen Y, Niemela P, Puhkakka M, Coley PD. 1986. River dynamics and the diversity of Amazon lowland
forest. Nature 332: 254–258.
Shelford VE. 1954. Some lower Mississippi valley floodplain biotic communities: their age and elevation. Ecology 35: 126–142.
Simons DB, Simons RK. 1987. Differences between gravel and sand bed rivers. In Sediment Transport in Gravel Bed Rivers, Thorne CR, Bath-
urst JC, Hey RD (eds.). John Wiley and Sons: Chichester.
Tabbacchi E, Planty-Tabbacchi AM, Salinas MJ, Decamps H. 1996. Landscape structure and diversity in riparian plant communities: a long-
itudinal comparitive study. Regulated Rivers: Research and Management 12: 367–390.
Tilman D, Pacala S. 1993. The maintenance of species richness in plant communities. In Species Diversity in Ecological Communities: Histor-
ical and Geographical Perspectives, Ricklefs RE, Schluter D (eds.). University of Chicago Press: London; 13–25.
Vandersman AJM, Joosten NN, Blom CWPM. 1993. Flooding regimes and life-history characteristics of short-lived species in river forelands.
Journal of Ecology 81: 121–130.
Van Eck WHJM, van de Steeg HM, Blom CWPM, de Kroon H. 2005. Recruitment limitation along disturbance gradients in river floodplains.
Journal of Vegetation Science 16: 103–110.
Vervuren PJA, Blom CWPM, de Kroon H. 2003. Extreme flooding events on the Rhine and the survival and distribution of riparian plant species.
Journal of Ecology 91: 135–146.
Ward JV. 1998. Riverine landscapes: biodiversity patterns, diversity regimes and aquatic conservation. Biological Conservation 83: 269–278.
Q6
Ward JV, Tockner K. 2001. Biodiversity: towards a unifying theme for river ecology. Freshwater Biology 46: 807–819.
Ward JV, Tockner K, Schiemer F. 1999. Biodiversity of floodplain river ecosystems: ecotones and connectivity. Regulated Rivers: Research and
Management 15: 125–139.
Ward JV, Tockner K, Uehlinger U, Malard F. 2001. Understanding natural patterns and processes in river corridors as the basis for effective
restoration. Regulated Rivers: Research and Management 17: 311–323.
Willby NJ, Pulford ID, Flowers TH. 2001. Tissue nutrient signatures predict herbaceous-wetland community responses to nutrient availability.
New Phytologist 152: 463–481.
Winterbottom SJ. 2000. Medium and short-term channel planform changes on the Rivers Tay and Tummel. Geomorphology 34: 195–208.
Wondzell SM, Cunningham GL, Bachelet D. 1996. Relationships between landforms, geomorphic process and plant communities on a
watershed in the Northern Chihuahuan Desert. Landscape Ecology 11: 351–362.
Q6
16 D. GIL VEAR AND N. WILLBY
Copyright #2005 John Wiley & Sons, Ltd. River Res. Applic. 21: 1–18 (2005)
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Appendix I: Summary of vegetation composition of principal habitat clusters on Ballinluig Island
C1 C2 C3 C4 C5 C1 C2 C3 C4 C5
Acer pseudoplatanus II 1 I 6 II 1 I 4 III 2 Lupinus polyphyllus I2 I7 II6 II4
Achillea millefolium I 2 I 3 III 2 Luzula sylvatica I1 I1 I2
Achillea ptarmica I1 I1 Lythrum salicaria I4
Aegopodagria podagraria I 5 II 4 Mentha aquatica I 1 II 3
Agrostis canina II 3 II 4 III 4 III 6 II 3 Mentha arvensis I 2 I 1 II 2
Agrostis capillaris I 2 I 3 III 3 IV 3 Mercurialis perennis I3
Agrostis stolonifera II 2 I 2 II 5 Montia fontana I1
Aira praecox I2 Mycelis muralis I1 I 2
Alnus glutinosa II 7 Myosotis scorpioides I 1 II 6 III 5 III 2
Anthoxanthum odoratum III 5 II 4 I 6 Ononis repens I 2 III 7 II 4 II 3 II 1
Anthriscus sylvestris I2 I2 I2 II2 Oxyria digyna I3
Anthyllis vulneraria II 5 Phalaris arundinacea I 2 I 2 II 8 V 3
Arrhenatherum elatius III 4 IV 6 II 3 I 4 II 3 Pinus sylvestris I 3 II 2 I 3
Betula pubescens I2 I 4 Plantago lanceolata II 2 II 5 III 3 III 4 IV 3
Bromus ramosus I2 I2 I1 Poa annua I2 II2 II2 I 4
Bryum pseudotriquetrum I 3 IV 5 I 3 I 5 Polygonum aviculare I1 I 1
Bryum sp I 6 I 3 I 6 III 6 Polygonum persicaria II 2
Callitriche hamulata I 2 I 2 II 2 Polytricum commune I2
Campanula rotundifolia I2 I2 I1 II2 Polytricum juniperinum III 6 I 2 II 2
Cardamine pratensis I3 Potentilla anserina I 6 I 3 III 3
Carex rostrata II 3 Prunella vulgaris I2 I4 II3 II3
Centaurea nigra I 5 II 1 II 3 III 5 IV 6 Racomitrium uliginosum I6
Cerastium fontanum II 3 I 1 Ranunculus flammula I 2 II 2
Chamaenerion angustifoilium II 2 Ranunculus repens II 5 I 2 II 2 II 2 III 3
Circaea alpina I4 Rhinanthus minor I2 I3 II3
Cirsium arvense I 3 I 2 II 3 Rhytidiadelphus squarrosus I3
Cochlearia officinalis I1 I2 I2 Rorippa palustris III 1 I 5 II 3 II 2
Crepis paludosa I 1 II 3 I 1 Rorippa sylvestris I2 I 2
Dactylis glomerata I 1 I 3 II 2 II 4 II 2 Rosa canina I3 I4 II3 II2
Danthonia decumbens I3 Rubus fruticosus I2 I1 I3 III3
Deschampsia caespitosa I2 I2 I3 Rubus idaeus I5
Digitalis purpurea I 1 II 2 I 1 I 2 Rumex acetosa I2 I2 I3 I2 II2
Eleocharis palustris II 4 Rumex acetosella I 4 II 3 II 5 II 2
Elymus caninus III 3 II 3 II 2 Rumex crispus I2
Epilobium montanum I1 I1 I1 Rumex obtusifolius I 1 II 2
Equisetum arvense I 3 II 3 Sagina procumbens I2
Festuca arundinacea I 3 III 6 III 7 Salix doniana?I7
Festuca ovina I 2 III 3 II 3 Salix aurita I8
Festuca rubra I 3 II 4 III 4 Salix cinerea I4 I7 II8 II2
Fillago minima I3 Sarothamnus scoparius III 6 IV 8 III 5 III 8
Fontinalis antipyretica I4 I 3 Schistidium alpicola I3 I1
Galeopsis tetrahit I1 II3 II2 I1 Scirpus sylvatica I 2 II 7
Galium boreale I5 Scrophularia nodosa I 2 II 2
Galium palustre I1 I4 Senecio jacobaea I1 I1 I2 II2 II3
Galium verum I 3 II 2 Senecio viscosa I 1 II 2 I 2 I 2
Geranium robertianum I2 I1 I2 Senecio vulgaris II 1 I 1
Geum rivale I2 I2 Silene maritimus V 7 V 6 V 6 III 4
Glyceria fluitans I2 Solidago virgaurea I2 I1 I3 II1
Heracleum sphondylium I1 I2 I2 Spergularia rubra I1
Hesperis matrionalis I1 Stellaria alsine I1 I2 I1 I1
Holcus lanatus II 2 III 5 II 4 I 5 Stellaria graminea I2 I2
Holcus mollis I2 I1 I2 II6 II5 Stellaria media I 2 II 3 I 1 I 2
Hypericum pulchrum I4 I2 I1 I2 III3 Succisa pratensis I1 I2 II3 II2
Impatiens glandulifera II 2 V 6 IV 3 III 2 II 2 Symphytum officinale I1 I2 II3 II3
Juncus acutiflorus II 3 Taraxacum agg IV 3 II 3 I 2 I 2
Juncus articulatus I 4 I 2 II 4 Teesdalia nudicaulis II 2 I 2
Leontodon autumnalis I 2 II 2 II 2 II 3 Teucrium scorodonia I 1 II 3 I 2 I 4
Continues
CONTROLS ON GRAVEL BAR VEGETATION
Q1
17
Copyright #2005 John Wiley & Sons, Ltd. River Res. Applic. 21: 1–18 (2005)
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Appendix I: Continued
C1 C2 C3 C4 C5 C1 C2 C3 C4 C5
Leontodon hispidus I 2 I 2 I 2 II 2 Thymus druceii I7 I3
Lepidium heterophyllum I3 II2 I2 I3 II2 Trifolium repens II 3 I 2 I 2
Leucanthemum vulgaris I 2 II 4 I 1 II 4 II 1 Tussilago farfara I2 I2 II4 II8
Linaria repens I4 Ulex europaeus I1 II3 II6
Lolium perenne I1 I1 I1 I2 II2 Veronica chamaedrys I 1 II 1
Lotus corniculatus II 2 I 1 Viccia cracca I3 I2
Lotus pedunculatus I7 Viola riviniana I 1 II 2 II 2 II 2 II 2
Note: values refer to frequency of occurrence within samples of each cluster (I20%, II ¼20–40%, III ¼40–60%, IV ¼60–80%, V >80%)
and maximum cover recorded within each cluster (1 0.1%, 2 ¼0.1–1%, 3 ¼1–2.5%, 4 ¼2.5–5%, 5 ¼5–10%, 6 ¼10–25%, 7 ¼25–50%,
8¼50–75%). In the interests of brevity species found only in a single sample with a cover of <1% have been excluded.
18 D. GIL VEAR AND N. WILLBY
Copyright #2005 John Wiley & Sons, Ltd. River Res. Applic. 21: 1–18 (2005)
UNCORRECTED PROOFS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Author Query Form (RRA/917)
Special Instructions: Author please write responses to queries directly on Galley proofs
and then fax back. Alternatively please list responses in an e-mail.
Q1: Author: Please check the suggested short title.
Q2: Author: Please check the change made in the spelling against the reference list.
Q3: Author: Please provide full name of the journal.
Q4: Author: Please update this information.
Q5: Author: Please check the page range.
Q6: Author: Reference not cited in text. Please check.
Q7: Author: There is a mismatch in hardcopy and e-file. So, we followed the hardcopy for
Fig 1, 5, and 6. and e-file for Fig 2, 3 and 4. Please check.
