HYDROMORPHOLOGICAL CONDITIONS, POTENTIAL FISH HABITATS AND THE FISH COMMUNITY IN A MOUNTAIN RIVER SUBJECTED TO VARIABLE HUMAN IMPACTS, THE CZARNY DUNAJEC, POLISH CARPATHIANS B. WYZ ˙ GA, a * A. AMIROWICZ, a A. RADECKI-PAWLIK b and J. ZAWIEJSKA c a Polish Academy of Sciences, Institute of Nature Conservation, al. Mickiewicza 33, 31-120, Krako ´w, Poland b Department of Water Engineering, Agricultural University, al. Mickiewicza 24/28, 30-059 Krako ´w, Poland c Institute of Geography, Pedagogical University, ul. Podchora˛z ˙ych 2, 30-084 Krako ´w, Poland ABSTRACT The Czarny Dunajec River, Polish Carpathians, has been considerably modified by channelization and gravel mining-induced channel incision and it varies in morphology from a single-thread, incised or regulated channel to an unmanaged, multi-thread channel. For twelve cross-sections with between 1 and 4 flow threads, the abundance and diversity of fish fauna were determined by electrofishing and compared with an assessment of hydromorphological river quality and the variation in flow depth, velocity and bed material size. Hydromorphological quality of the river varied between high-status conditions (Class 1) in unmanaged, multi-thread cross-sections with varying proportions of islands and Class 4 in channelized cross-sections. The increased number of low-flow channels in a cross-section was associated with a larger aggregated width of low-flow channels and greater variation in flow depth, velocity and bed material size. Single-thread cross-sections hosted only 2 fish species and 13 individuals on average, whereas 3–4 species and 82 individuals on average were recorded in cross-sections with four low-flow channels. Regression analysis indicated that both the number of fish species and individuals increased linearly with increasing variation in depth within a cross-section and exponentially with improving hydromorphological river quality, while they were unrelated to flow width, suggesting that it is the increase in variability of habitat conditions rather than simple habitat enlargement, that supported the increased abundance and diversity of fish fauna in the multi-thread cross-sections. This study shows that the simplification of flow pattern and the resultant degradation of hydromorphological river quality, caused by human impacts, is reflected in remarkable impoverishment of fish communities and that recovery of these communities will require an increase in morphological complexity of the river. Copyright # 2009 John Wiley & Sons, Ltd. key words: mountain river; human impact; hydromorphological river quality; fish habitats; fish fauna Received 19 December 2008; Accepted 23 December 2008 INTRODUCTION During the 20th century most reaches of Polish Carpathian rivers were greatly modified by human activities. Intense channelization works resulted in considerable channel narrowing, accompanied by significant channel shortening in the lower reaches of rivers and the replacement of former, multi-thread channels by single-thread, artificial ones in the middle and upper reaches (Wyz ˙ga, 2001a, 2008; Bojarski et al., 2005). In the 1940s–1960s huge amounts of gravel were mined from some Carpathian rivers (Rinaldi et al., 2005) and in the subsequent decades illegal exploitation of bed material from mountain stream and river channels was a common practice (Wyz ˙ga, 2001a; Radecki-Pawlik, 2002). The increase in transport capacity of Carpathian rivers caused by channelization and also removal of bed material resulted in rapid bed degradation (Wyz ˙ga, 2008), which was strengthened in the second half of the century by a reduction in sediment delivery resulting from an increase in forest cover in the montane parts of catchments (Lach and Wyz ˙ga, 2002). As a result, Carpathian rivers experienced channel incision of between 0.5 and 3.8 m during the 20th century, with rates of incision being particularly high in many river sections during the second half of the century. Incision has increased the flow capacity of channels and so has reduced the frequency and lateral extent of inundation of valley floors, inducing a marked increase in flood hazard to RIVER RESEARCH AND APPLICATIONS River. Res. Applic. 25: 517–536 (2009) Published online 4 February 2009 in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/rra.1237 *Correspondence to: B. Wyz ˙ga, Polish Academy of Sciences, Institute of Nature Conservation, al. Mickiewicza 33, 31-120 Krako ´w, Poland. E-mail: [email protected]Copyright # 2009 John Wiley & Sons, Ltd.
20
Embed
Hydromorphological conditions, potential fish habitats and the fish community in a mountain river subjected to variable human impacts, the Czarny Dunajec, Polish Carpathians
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
RIVER RESEARCH AND APPLICATIONS
River. Res. Applic. 25: 517–536 (2009)
Published online 4 February 2009 in Wiley InterScience
HYDROMORPHOLOGICAL CONDITIONS, POTENTIAL FISH HABITATS ANDTHE FISH COMMUNITY IN A MOUNTAIN RIVER SUBJECTED TO VARIABLE
HUMAN IMPACTS, THE CZARNY DUNAJEC, POLISH CARPATHIANS
B. WYZGA,a* A. AMIROWICZ,a A. RADECKI-PAWLIKb and J. ZAWIEJSKAc
a Polish Academy of Sciences, Institute of Nature Conservation, al. Mickiewicza 33, 31-120, Krakow, Polandb Department of Water Engineering, Agricultural University, al. Mickiewicza 24/28, 30-059 Krakow, Poland
c Institute of Geography, Pedagogical University, ul. Podchora zych 2, 30-084 Krakow, Poland
The Czarny Dunajec River, Polish Carpathians, has been considerably modified by channelization and gravel mining-inducedchannel incision and it varies in morphology from a single-thread, incised or regulated channel to an unmanaged, multi-threadchannel. For twelve cross-sections with between 1 and 4 flow threads, the abundance and diversity of fish fauna were determinedby electrofishing and compared with an assessment of hydromorphological river quality and the variation in flow depth, velocityand bed material size. Hydromorphological quality of the river varied between high-status conditions (Class 1) in unmanaged,multi-thread cross-sections with varying proportions of islands and Class 4 in channelized cross-sections. The increased numberof low-flow channels in a cross-section was associated with a larger aggregated width of low-flow channels and greater variationin flow depth, velocity and bed material size. Single-thread cross-sections hosted only 2 fish species and 13 individuals onaverage, whereas 3–4 species and 82 individuals on average were recorded in cross-sections with four low-flow channels.Regression analysis indicated that both the number of fish species and individuals increased linearly with increasing variation indepth within a cross-section and exponentially with improving hydromorphological river quality, while they were unrelated toflow width, suggesting that it is the increase in variability of habitat conditions rather than simple habitat enlargement, thatsupported the increased abundance and diversity of fish fauna in the multi-thread cross-sections. This study shows that thesimplification of flow pattern and the resultant degradation of hydromorphological river quality, caused by human impacts, isreflected in remarkable impoverishment of fish communities and that recovery of these communities will require an increase inmorphological complexity of the river. Copyright # 2009 John Wiley & Sons, Ltd.
key words: mountain river; human impact; hydromorphological river quality; fish habitats; fish fauna
Received 19 December 2008; Accepted 23 December 2008
INTRODUCTION
During the 20th century most reaches of Polish Carpathian rivers were greatly modified by human activities. Intense
channelization works resulted in considerable channel narrowing, accompanied by significant channel shortening
in the lower reaches of rivers and the replacement of former, multi-thread channels by single-thread, artificial ones
in the middle and upper reaches (Wyzga, 2001a, 2008; Bojarski et al., 2005). In the 1940s–1960s huge amounts of
gravel were mined from some Carpathian rivers (Rinaldi et al., 2005) and in the subsequent decades illegal
exploitation of bed material from mountain stream and river channels was a common practice (Wyzga, 2001a;
Radecki-Pawlik, 2002). The increase in transport capacity of Carpathian rivers caused by channelization and also
removal of bed material resulted in rapid bed degradation (Wyzga, 2008), which was strengthened in the second
half of the century by a reduction in sediment delivery resulting from an increase in forest cover in the montane
parts of catchments (Lach and Wyzga, 2002). As a result, Carpathian rivers experienced channel incision of
between 0.5 and 3.8m during the 20th century, with rates of incision being particularly high in many river sections
during the second half of the century. Incision has increased the flow capacity of channels and so has reduced the
frequency and lateral extent of inundation of valley floors, inducing a marked increase in flood hazard to
*Correspondence to: B. Wyzga, Polish Academy of Sciences, Institute of Nature Conservation, al. Mickiewicza 33, 31-120 Krakow, Poland.E-mail: [email protected]
Copyright # 2009 John Wiley & Sons, Ltd.
518 B. WYZGA ET AL.
downstream reaches (Wyzga, 1997, 2008) and a reduction in the potential of Carpathian floodplains to store
sediment (Wyzga, 2001b, 2008). With these and many other detrimental effects of channel incision becoming
apparent at a local scale, concrete weirs were constructed in many Carpathian stream and river channels to arrest
bed degradation and reduce bedload transport (Bednarczyk et al., 2003). This construction has disrupted the
continuity of the watercourses for fish (Bojarski et al., 2005; Wisniewolski, 2005).
In the 20th century, many European mountain and piedmont rivers were subjected to similar human disturbances
and the resultant loss of vertical channel stability (e.g. Bravard et al., 1997; Surian and Rinaldi, 2003) and decrease
in the biodiversity of aquatic and riparian ecosystems (e.g. Roux et al., 1989; Muhar et al., 2008). With increasing
understanding of the adverse effects of human activities in river channels, restoration measures are being
undertaken widely to improve their geomorphological and ecological conditions (Habersack and Piegay, 2008),
strengthened by the Water Framework Directive of the European Union (Directive 2000/60/EC), which requires
that rivers will attain good ecological status by 2015.
To restore good ecological status of watercourses, it is essential to determine the relations which exist between
the structure of river biocoenoses and the hydromorphological characteristics of the habitats. Such relationships
support assessment of the degree to which the good quality of river ecosystems may be re-established through the
improvement of physical characteristics as well as the flow and sediment regime of the watercourse (i.e. its
hydromorphological quality), or whether ecosystem degradation has resulted from other factors such as a decrease
in water quality. Moreover, determining gradients in the biodiversity of riverine communities in relation to the scale
of habitat modification with respect to reference conditions may support identification of the range of restoration
measures which need to be undertaken.
This paper aims to determine relationships between the abundance and diversity of fish fauna and
hydromorphological conditions in a number of cross-sections of the Czarny Dunajec draining the Polish
Carpathians. Among the groups of biota used as indicators of the ecological status of watercourses (fish,
macroinvertebrate and diatom communities), fish have the longest life cycle and so are likely to be more sensitive to
long-term changes in physical characteristics of their habitats than to short-term changes in water quality. Thus, the
status and condition of fish fauna are a useful monitoring tool to identify degradation in the ecological integrity of
rivers caused by human-induced modifications of hydromorphological conditions (Schiemer, 2000).
STUDY AREA
The gravel-bed Czarny Dunajec (Figure 1) constitutes the upper part of the Dunajec, the second largest river of the
Polish Carpathians. The Czarny Dunajec rises at about 1500m a.s.l. in the high-mountain Tatra massif. In the Tatra
Mountains foreland, it flows for a length of about 38 km to its confluencewith the Biały Dunajec River at an altitude
of 578m (Figure 1B).
During the second half of the 20th century, the river channel was affected by considerable, spatially variable
human modification (Krzemien, 2003; Zawiejska and Krzemien, 2004; Zawiejska and Wyzga, submitted to press).
In the 1950s–1960s intense gravel mining from the river bed occurred at several locations (Dudziak, 1965),
followed in subsequent decades by widespread, illegal extraction of larger cobbles (Krzemien, 2003). These
activities induced up to 3.5m of channel incision and transformation from an alluvial to bedrock bed in many
sections of the river within the Gubałowka Hills (Zawiejska and Wyzga, submitted to press). In the 1960s–1990s, a
7 km long stretch in the middle river course was progressively channelized using 0.7–2.1m high concrete drop
structures, which disrupted fish passage, and the former multi-thread channel was replaced by a nearly straight,
single-thread one. Immediately downstream of the channelized stretch, the river remains unmanaged over a length
of about 4 km, with its channel pattern varying from braided through island-braided to heavily island-braided.
Further downstream, throughout the lower course of the river, the channel was considerably straightened and
narrowed by channelization works, which were completed by the 1980s, but the channel gradient was not reduced
by construction of drop structures.
The study reported in this paper was performed in a 17 km long reach in the middle river course, along which the
catchment area increases relatively little and the river receives no major tributaries (Figure 1B). This reach
comprises a deeply incised channel in its upper part, a regulated channel with drop structures in its middle part, an
Figure 1. (A) Location of the Czarny Dunajec River in relation to physiogeographic regions of southern Poland. (B) Drainage network andphysiography of the Czarny Dunajec catchment and detailed setting of the investigated sections of the river: (1) high mountains; (2) mountains ofintermediate and low height; (3) foothills; (4) intramontane and submontane depressions; (5) the Czarny Dunajec catchment to the beginning ofthe study reach; (6) catchment area increment along the study reach; (7) boundary of the Czarny Dunajec catchment; (8) boundaries ofphysiogeographic units; (9) river cross-sections investigated; (10) flow-gauging stations; PKB, Pieniny Klippen Belt; STT, Sub-Tatran Trough
HYDROMORPHOLOGY AND FISH COMMUNITY IN MOUNTAIN RIVER 519
unmanaged channel in the lower part and again a regulated channel at its downstream end. The reach is typified by
considerable variation in river morphology (Wyzga and Zawiejska, 2005) including single- and multi-thread
sections, sections with alluvial and bedrock boundaries and with incised and vertically stable beds. Together with
differences in channel management, these contrasting channel conditions are associated with high variability in
physical habitat conditions and are likely to be reflected in differences between local river biocoenoses. Indeed,
considerable variation in the ecological status of the Czarny Dunajec in its middle course was revealed by combined
evaluation of both biotic and abiotic characteristics of the river performed in the 1990s (Kulesza, 1998). Since
channel incision in the upper part of the study reach has continued, the present downstream variations in
hydromorphological conditions within the reach may be even larger than were observed in the 1990s.
STUDY METHODS
Investigations were conducted at twelve river cross-sections, representing the range of hydromorphological
conditions present in the study reach of the Czarny Dunajec. Hydromorphological river quality in the cross-sections
was assessed through scoring of ten groups of features of the channel, river banks, riparian zone and floodplain
(Table I), mainly according to their specification in the European Standard EN-14614 (CEN, 2004), although two
additional features were added (Table I). First, an appraisal of in-channel, aquatic vegetation was replaced by
separate appraisal of aquatic vegetation and island vegetation within the river’s active zone. This reflected a
considerable role played by island vegetation in the Czarny Dunajec and the observation that the island vegetation
may significantly differ in naturalness from the vegetation in the riparian zone. Second, in the assessment of
modifications to river flow, we completed an appraisal of discharge regime and flow hydraulics by evaluating the
degree of connectivity between river channel and hyporheic zone. While the significance of water exchange
between river channel and hyporheic zone for river biota is well recognized (e.g. Baxter and Hauer, 2000), its
occurrence may be drastically reduced or even eliminated by channel incision which induces a change from an
alluvial to bedrock river bed.
A reference state for the river was not derived from historical information because its previous braided state,
particularly in the 19th century, was partly a response to intense agricultural and pastoral activities. Instead, we
considered reference hydromorphological conditions as those that would exist under present environmental
Table I. Assessment categories and features of the channel, banks, riparian zone and floodplain of the Czarny Dunajec assessedto evaluate the hydromorphological quality of the river according to the European Standard EN-14614 (CEN, 2004)
No. Assessment category Assessed feature
Channel1 Channel geometry Channel planform
Channel cross-section and longitudinal profile2 Substrate Artificial/natural bed substrate
Degree of modification of substrate material3 In-river vegetation and organic debris Aquatic vegetation and island vegetation within
the river’s active zone�
Organic material (leaves, woody debris)4 Erosion/deposition features Presence of erosional and depositional channel forms5 Flow Modification of natural flow hydraulics by engineering
works and structuresModification of natural flow regimeDegree of connectivity between river channel andhyporheic zone�
6 Longitudinal river continuity Impact of engineering structures on longitudinal rivercontinuity—sediment transport and biota migration
Banks and Riparian Zone7 Bank structure Modification of river banks (material, profile, height)8 Riparian zone Vegetation and land use in the riparian zoneFloodplain9 Land use and associated features in the river corridor Land use in the river corridor/floodplain
Presence of remnant channels, oxbows and mires withinthe floodplain
10 Lateral connectivity and channel migration Degree of lateral connectivity of river and floodplain,and continuity of floodplain along the riverConstraint on lateral channel movement
�feature modified or added to the list indicated in the European Standard EN-14614.
HYDROMORPHOLOGY AND FISH COMMUNITY IN MOUNTAIN RIVER 521
conditions in the catchment but without human influence on the channel, riparian zone and floodplain of the river.
The assessment was performed simultaneously by four specialists in fluvial geomorphology, river engineering and
hydrobiology following field inspection of the cross-sections, the analysis of river change in their vicinity over the
past few decades as well as the presentation of diagrams of the cross-sections and river appearance in particular
cross-sections on orthophotos and ground photos (Figure 2). Each assessed category was scored on the scale from
1 (for near-natural conditions) to 5 (for extremely modified conditions). The aggregated score, averaged from the
scores of the four specialists, allowed each of the cross-sections to be associated with a particular class of
hydromorphological quality.
Figure 2. Examples of the sources of information used to evaluate hydromorphological quality of the Czarny Dunajec River in the surveyedcross-sections, illustrated for cross-section D with single-thread, regulated channel and cross-section K with multi-thread, partly reinforced
channel. This figure is available in colour online at www.interscience.wiley.com/journal/rra
The investigated cross-sections were located to run consistently across pools, so as to examine differences
between particular types of river morphology and channel management rather than those related to pool-riffle
sequences in the river. Detailed levelling of the cross-sections was performed during baseflow conditions and water
depth, mean flow velocity and mean grain size of surface bed material were determined at 1m intervals within the
low-flow channel(s). For gravelly sediments, transect sampling (which yields results equivalent to bulk sieve
analysis; Diplas and Sutherland, 1988; ISO 9195, 1992) was used, with 15 particles collected from the bed at each
site; this number of particles was chosen to avoid extension of the sampling beyond the area characterized by
hydraulic measurements. Subsequently, the distribution of the ‘b’ axis diameters of collected particles was
determined and their mean size was calculated as an average of the diameter of the 3rd, 8th and 13th grain in the
sequence (reflecting the 20th, 50th and 80th percentiles of the grain size distribution), this being the closest
available approximation of the formula of Folk and Ward (1957). The grain size distribution of sandy and silty
sediments was established in a laboratory using sieving or hydrometer analyses and their mean grain size was
calculated from the same percentiles of the distribution. For each cross-section, means and coefficients of variation
of water depth, mean flow velocity and mean grain size of surface bed material were then calculated.
Species composition of fish communities in the investigated cross-sections was estimated on the basis of single
electrofishings conducted on 4 September 2006. Fishes were caught in a 10m wide strip along each cross-section
using the backpack electroshocker (DEKA-Lord). Number and approximate total lengths of fish specimens caught
in particular low-flow channels were recorded. Two age categories were distinguished: juveniles (YOY) as well as
subadult and adult fishes (1þ age class and older). Specimens were assigned to these categories on the basis of their
total length, taking the length of 10 cm (brown trout) or 5 cm (other species) as the size limit.
RESULTS
Hydromorphological river quality
Evaluation of the hydromorphological quality of the river indicated considerable variation in this parameter
among the surveyed cross-sections, with the values ranging between 1.08 and 3.96 (Table II). For all cross-sections
with four low-flow channels, the river was classified as representing high status (reference) conditions. Cross-
sections I and J, showing heavily island-braided morphology, were considered to be only slightly modified by
human activity (scores 1.08 and 1.26), while a somewhat greater degree of modification, largely as a result of
artificial reinforcement of one of the channel banks, typifiedmulti-thread cross-sections H and Kwith moderate and
scarce occurrence of islands, respectively (scores 1.57 and 1.69). Single-thread cross-sections with regulated
Table II. Averages of the scores given by four evaluators for particular assessment categories and average scores of thehydromorphological river quality in the investigated cross-sections of the Czarny Dunajec. The average scores for particularcross-sections are shown against hydromorphological quality classes
Evaluated cross-section A B C D E F G H I J K LNumber of flow threads I III I I I I II IV IV IV IV I
HYDROMORPHOLOGY AND FISH COMMUNITY IN MOUNTAIN RIVER 523
channels (D–F and L) were scored between 3.54 and 3.96, hence falling into Class 4 of hydromorphological quality.
This reflected radical modification of channel geometry in the cross-sections (Figure 2—cross-section D), absence
of erosional and depositional forms, disturbance of fish migration by weirs, loss of hydraulic connectivity of the
river with its floodplain and lateral channel stabilization (Table II). Finally, unmanaged cross-sections A–C (two
single-thread cross-sections and one cross-section with three low-flow channels) with deeply incised channel as
well as cross-section G with two low-flow channels and partly reinforced banks were considered to represent
hydromorphological quality Class 2, as a result of modest deviation from reference conditions for a number of
features that were evaluated (Table II).
As the evaluation of hydromorphological quality of the river was based on expert judgement, it is interesting to
see the degree of agreement between scores assigned to particular cross-sections by the four specialists and the high
similarity in the opinions of experts for most of the evaluated cross-sections (Figure 3). The opinions differed by
about 1 quality class for only three cross-sections: cross-section C, in which the river had incised by about 3m over
the second half of the 20th century; cross-section F with narrow, regulated channel and moderately wide cross-
section G with partly reinforced banks. For each of these cross-sections, better opinions on a number of evaluated
features were expressed by a hydrobiologist than by both fluvial geomorphologists who were more aware of
human-induced changes to the channel morphology and their impact on hydraulic and sedimentary characteristics
of the river (increase in flow velocity, simplification of the flow pattern, coarsening of bed material) in these
locations (cf. Wyzga et al., 2009).
Physical parameters of riverine habitats
Measurements of hydraulic parameters were performed at discharges varying between 3.15 and 3.74m3 s�1
among the surveyed cross-sections. This variation reflected run-off variability during the measurement campaign
and the specific hydrographic setting of the study reach (Figure 1). No significant systematic change in the
measured discharge was found either with the position of a particular cross-section in the downstream sequence
(linear regression, p¼ 0.48) or with the distance of a cross-section from the beginning of the study reach (p¼ 0.36).
Therefore, it was possible to consider that the data from all cross-sections represented similar flow conditions and
Figure 3. Scores given by four evaluators for the hydromorphological river quality in the investigated cross-sections of the Czarny Dunajec andthe average of the four estimates. The number of flow threads in each cross-section is also indicated
thus to analyse whether the physical parameters of fish habitats in the river were dependent on the complexity of
flow network. Multi-thread cross-sections were typified by significantly greater (Mann–Whitney test, p¼ 0.03)
aggregated width of low-flow channels than single-thread cross-sections (Figure 4), with widths of 19.8m on
average in single-thread cross-sections and 34.1m in cross-sections with four low-flow channels (Figure 5). An
Figure 4. Examples of cross-sectional morphology of the Czarny Dunajec in channelized river sections (upper) and an unmanaged section(lower). For low-flow channels, mean grain size of the sediment on the bed surface and mean flow velocity are indicated at 1m intervals. The
scale for velocity commences at the water surface for each low-flow channel
Figure 5. Scatter plots of physical characteristics of the Czarny Dunajec versus the number of flow threads in the investigated cross-sections.The significance of the difference of the parameter means between the cross-sections with one and four low-flow channels were determined by
Mann–Whitney test. The dashed lines link the parameter means between the two groups of cross-sections
Figure 6. Scatter plots for the pairs of physical characteristics of the Czarny Dunajec measured at all sampling sites in the investigated single-thread cross-sections (left diagrams) and multi-thread cross-sections (right diagrams) of the river
HYDROMORPHOLOGY AND FISH COMMUNITY IN MOUNTAIN RIVER 527
3–4 species and 36–119 individuals (82 on average), with the differences in the number of both species and
specimens being statistically significant (Mann–Whitney test, p¼ 0.01 in both cases). Multi-thread cross-sections
also hosted numerous juveniles (mean: 166) in comparison with very low numbers of such fishes in single-thread
cross-sections (only two specimens on average) (Figure 7). Moreover, these two types of cross-sections differed in
the proportion of juvenile and older specimens (x2 test, p< 0.0001). In the multi-thread cross-sections, juveniles
were 2.7 times more abundant than subadult and adult individuals (998 and 375), whereas in the single-channel
cross-sections they were 6.5 times less numerous (12 and 78).
However, these relations did not follow a simple spatial pattern within the study reach (Figures 7 and 8).
The greatest numbers of both juvenile and older specimens occurred in cross-sections I and J which had a heavily
island-braided morphology and were located in the central part of a 4 km long, multi-thread channel section.
Despite the same number of low-flow channels, cross-sections H and K, which supported a moderate to scarce
occurrence of islands and were located close to the ends of the multi-thread channel section, hosted less abundant
Figure 7. Species composition of fish assemblages recorded during single electrofishing surveys within low-flow channels of the investigatedcross-sections of the Czarny Dunajec. This figure is available in colour online at www.interscience.wiley.com/journal/rra
528 B. WYZGA ET AL.
juvenile and older fishes. Two-species fish assemblages with relatively low numbers of individuals were found in
single-thread cross-sections located both upstream and downstream of the unmanaged, multi-thread channel
section. A similar assemblage with only two species and quite low number of individuals was also recorded in
multi-thread cross-section B, which had a bar-braided morphology and was located in the deeply incised, upper part
of the study reach.
Relationships between the abundance and diversity of fish and explanatory variables
With the detailed recognition of a set of physical parameters for the surveyed cross-sections, it was possible to
verify whether they can explain the observed variability in the abundance and species diversity of fish among the
Figure 8. Comparison of the number of low-flow channels, the assessment of hydromorphological quality and the results of electrofishingcarried out in 12 cross-sections of the Czarny Dunajec River: St, Salmo trutta L.; Cp, Cottus poecilopus Heckel; Pp, Phoxinus phoxinus (L.);Bb, Barbatula barbatula (L.); n, total number of subadults and adults (�1 year old) caught; dashed line indicates the occurrence of
juveniles (YOY) only. This figure is available in colour online at www.interscience.wiley.com/journal/rra
HYDROMORPHOLOGY AND FISH COMMUNITY IN MOUNTAIN RIVER 529
cross-sections. Results from regression analysis indicated that the number of fish species as well as subadult and
adult individuals increased linearly with increasing variation in water depth in a cross-section (Figure 9, Table III).
The number of individuals also increased with decreasing grain size of material on the channel bed and with
increasing variation of grain size of the material in a cross-section. Other physical parameters under consideration
had no significant effect on fish communities (Table III).
The species diversity of fish communities and fish abundance in the surveyed cross-sections were greater in the
cross-sections of better hydromorphological quality (Figure 8). Regression analysis showed that fish diversity and
abundance increase exponentially with improving hydromorphological river quality (i.e. decreasing quality scores)
(Figure 9, Table III). For fish abundance, this non-linear relationship was especially pronounced and the
greatest increase in the number of individuals occurred with a relatively small improvement in hydromorphological
quality of the river that accompanied the change from braided morphology with occasional islands to heavily
Figure 9. Scatter plots and estimated regression relationships between the number of fish species (left diagram) and the number of subadult andadult fish (right diagram) caught in the investigated cross-sections of the Czarny Dunajec, and the coefficient of variation of flow depth and
hydromorphological river quality in the cross-sections
530 B. WYZGA ET AL.
island-braided morphology. Importantly, the relationship with hydromorphological river quality explains a
considerably greater portion (88%) of the total variation in fish abundance than the regression models estimated for
single physical parameters of fish habitats (between 34% for the coefficient of grain size variation and 50% for the
coefficient of flow depth variation) (Table III).
Table III. Explanation and significance of regression models estimated between the number of fish species and the number ofsubadult and adult fishes caught in the investigated cross-sections of the Czarny Dunajec River, and potential explanatoryvariables. Relationships with p-values < 0.05 are indicated in bold
Independent variable Dependent variable Type ofregression