Biological and hydromorphological integrity of the small urban stream

This article was originally published in a journal published byElsevier, and the attached copy is provided by Elsevier for the

author’s benefit and for the benefit of the author’s institution, fornon-commercial research and educational use including without

limitation use in instruction at your institution, sending it to specificcolleagues that you know, and providing a copy to your institution’s

administrator.

All other uses, reproduction and distribution, including withoutlimitation commercial reprints, selling or licensing copies or access,

or posting on open internet sites, your personal or institution’swebsite or repository, are prohibited. For exceptions, permission

may be sought for such use through Elsevier’s permissions site at:

http://www.elsevier.com/locate/permissionusematerial

Autho

r's

pers

onal

co

py

Biological and hydromorphological integrity of the small urban stream

Branka Tavzes *, Gorazd Urbanic, Mihael J. Toman

Biotechnical Faculty, Department of Biology, University of Ljubljana, Vecna pot 111, 1000 Ljubljana, Slovenia

Received 25 July 2005; accepted 6 July 2006Available online 10 October 2006

Abstract

Biological and hydromorphological integrity of five reaches of the small urban stream were assessed. Because macroinvertebrate com-munities respond to both organic pollution and habitat change, impacts of both measures can be hardly separated. In our study on theurbanized small stream, an impact of organic pollution was excluded as all five sampling sites were assessed as moderately polluted. Onthe other hand differences in morphological degradation of banks and channel of selected sites enabled us to relate hydromorphologicalstress and biotic metrics and taxa. Physical habitat quality was assessed using River habitat survey (RHS) methodology. A downstream–upstream gradient of physical habitat degradation was observed and related to the macroinvertebrate community characteristics. Sim-ilarity analyses and biotic metrics were calculated and correlated with results of the RHS analyses. Composition of the macroinvertebrateassemblages did not follow the longitudinal pattern of habitat modification observed by the RHS analysis. However, some metrics cor-responded well. Percentage of detritivores, percentage of Caenis luctuosa, number of individuals, percentage of EPT individuals were bestpredictors of changes in the physical habitat quality. However, the metric percentage of EPT individuals was negatively correlated to thehabitat degradation, what is in contradiction with results from studies of other authors.� 2006 Elsevier Ltd. All rights reserved.

Keywords: Urban stream; Macroinvertebrates; Habitat modification; Ecological quality; RHS

1. Introduction

Macroinvertebrate communities are widely used toassess the biological integrity of running waters (Beiselet al., 2000). They respond to both organic pollution andhabitat change (Buffagni et al., 2000; Beavan et al., 2001;Lorenz et al., 2004). In urban streams macroinvertebratesare often exposed to combination of both factors. Chan-nelisation is the most common habitat change in urbanareas. Ecological consequences of channelisation are thathabitat diversity and niche potential are reduced and thequality and function of the species are changed (Brookesand Gregory, 1998). The number of taxa appears to behigher in a heterogeneous environment, where habitatsare more varied and a higher number of taxa can poten-

tially find ecological niches (Beisel et al., 2000). Urbanizedlotic systems are often classified as having poor or verypoor biotic integrity as in a homogeneous environmentthere is usually a particular type of microhabitat that canshelter a large number of individuals of a specific set of spe-cies. Aggregation of very adapted species is thereforefavoured in a very homogeneous environment (Daviset al., 2003; Beisel et al., 2000). In organically polluted riv-ers Beavan et al. (2001) found biotic scores and beta diver-sity analysis to indicate that less modified sites supported aslightly higher quality invertebrate fauna than physicallymodified sites. Altered fluvial processes as a result ofurbanization and their influence upon channel geomor-phology may be a more fundamental determinant in thefunctioning of urban stream ecosystems than factorsrelated to toxicity or organic loading. Urban rivers arestraightened and banks reinforced in an attempt to channelhigh flows out of the area quickly and safely. The alteredfeatures of an urbanized stream system may impact the

1474-7065/$ - see front matter � 2006 Elsevier Ltd. All rights reserved.

doi:10.1016/j.pce.2006.07.009

* Corresponding author. Tel.: +386 1 423 33 88; fax: +386 1 257 33 90.E-mail addresses: [email protected] (B. Tavzes), gorazd.urba-

[email protected] (G. Urbanic).

www.elsevier.com/locate/pce

Physics and Chemistry of the Earth 31 (2006) 1062–1074

Autho

r's

pers

onal

co

py

bottom fauna directly through specific habitat changes orindirectly through the temporal reduction in the quantityof available food (Pedersen and Perkins, 1986). Urbanizedstreams though, seem to be preferred by less specializedfeeders due to inconsistency of food sources (as decayingleaves) that are rapidly transported downstream (Pedersenand Perkins, 1986).

In the presented study both the hydromorphological andthe biological integrity of the selected stream reaches of theGlinscica stream were assessed. In order to determine howchanges in hydromorphological integrity of the stream affectthe biological one the correlations were calculated betweenstream hydromorphological characteristics and the bioticmetrics based on benthic macroinvertebrate fauna. Glin-scica stream has been chosen for a case study system, whichexhibits modifications of the hydromorphological charac-teristics as well as changes in riparian and channel vegeta-tion, which progress downstream in a relatively shortdistance from a pristine to a heavily modified channel.

2. Description of the study site

The Glinscica stream, a tributary of the LjubljanicaRiver, lies on the western outskirts of Ljubljana in the cen-tre of Slovenia (Fig. 1).

Elevations of the stream range from 500 m above meansea level at the source to 289 m at the mouth. The relief ofthe Glinscica watershed is heterogeneous with steep head-water areas in the north and plains in the south. By extend-ing the paved impermeable urban areas on the plain partsof the watershed, the hydrology of the entire watershedhas changed dramatically in the last 20 years. Prior tointensive urbanization, these plain areas did not contributein the formation of peaks of runoff hydrographs. Theincrease in impervious surfaces (extension of building, traf-fic surfaces) brought about the augmentation of the runoffcoefficient. Furthermore, the drainage system caused a fur-ther decrease in the concentration time of rainfall runoff.Urban areas present 38% of the entire Glinscica watershed,i.e. 6.6 km2. The approximate value for the average runoffcoefficient, computed from the average annual rainfall(1376 mm) and average annual discharge in the Glinscica

stream (0.383 m3/s), is 0.58 (Rusjan et al., 2003). Besideschanges in the watershed and hydrology there are alsochanges in the stream hydromorphology. Modificationsinclude stream habitat characteristics as well as changesin riparian vegetation and progress downstream. In thelower stretch the stream has concrete channel and banksand flows through urban area (Tables 1 and 2).

3. Materials and methods

Five sampling sites of different habitat quality (Table 1)were chosen for biological and hydromorphological analy-ses (Fig. 1). Each sampling site was selected as a represen-tative of the certain state of the habitat degradation.Site GLIN 1, the uppermost, was located in the nearly pris-tine reach, whereas the other four in reaches affected bythe anthropogenic land-use disturbances, which progress

Fig. 1. Sketch of the study area with location of sampling sites on theGlinscica stream.

Table 1Location and description of sampling sites of the Glinscica stream

Sitecode

Location Distance tosource (km)

Elevation(m)

Streamorder

Bank/channel modifications Riparian vegetation

GLIN 1 Podutik 2.3 310 2 None Dense riparian vegetation(trees and shrub)

GLIN 2 Kosezebridge

4.6 302 2 Straightened channel Sporadic trees and shrub

GLIN 3 Brdnikovastreet

6.0 300 3 Straightened channel, bank reinforcement withstones

Grass

GLIN 4 PST bridge 6.2 299 3 First half trapezoidal reinforced banks, the otherhalf trapezoidal concrete channel

Grass

GLIN 5 Biotechnicalfaculty

7.1 298 3 Trapezoidal concrete channel Grass

B. Tavzes et al. / Physics and Chemistry of the Earth 31 (2006) 1062–1074 1063

Autho

r's

pers

onal

co

py

downstream. The whole selected section of the streambelongs to the same abiotic river type according to the Slo-venian river typology (Urbanic, 2005). This fulfils demandsof the water framework directive (EU, 2000) that type-spe-cific approach must be used for the assessment of the eco-logical integrity of the waters.

The RHS survey (Environment Agency, 2003) was con-ducted for each sampling site. Survey data were analysedusing the habitat quality assessment scoring system and hab-itat modification score rules version 3.3 (EnvironmentAgency, 2002). The output of the calculation gave us a totalhabitat quality assessment score (HQA) and habitat modifi-cation score (HMS). HMS gives us information about thedegree of alteration on the selected stream reaches. Accord-ing to the HMS scores sites were allocated in HMS classes.

Biological sampling was conducted in September 2004.At each sampling site one macroinvertebrate sample wastaken with a hand net (500 lm mesh size) using a time-lim-ited (3 min) multi-habitat kick-sampling method. The loca-tion of the macroinvertebrate sampling site has beencentred at the eight RHS spot-check (the third most down-stream transect). Collected organisms were sorted accord-ing to taxonomic groups, preserved in 70% ethanol andidentified to the species level where possible, otherwise tothe highest possible level. Taxa list and counts were madeand used in subsequent data analyse.

The differences in macroinvertebrate community struc-ture between the sites were assessed by the cluster analysisusing Bray–Curtis similarity index and detrended corre-spondence analysis (DCA). Both calculations were madeusing log transformed data. In addition, a series of macro-invertebrate metrics were calculated for each sampling site.These ranged from taxa richness measures (number of taxa,

number of Ephemeroptera, Plecoptera and Trichoptera(EPT) taxa) and diversity measure (Shannon–Wiener diver-sity index) to assemblage composition (log number of indi-viduals, percentage of EPT taxa, percentage of EPTindividuals, percentage of Caenis luctuosa) and functionalfeeding groups (% shredders, % grazers, % filterers, % detri-tivores, % predators, % parasites and % other). Functionalfeeding groups were assigned according to Moog (1995).

Organic pollution of the selected sampling sites wasassessed using a saprobic index (SI). The formula ofZelinka and Marvan (1961) and saprobic values and indic-ative weights of Moog (1995), Wegl (1983) and Urbanic(2004) were used for the index calculations. Iberian Biolog-ical Monitoring Working Party score (IBMWP) was calcu-lated as well, because the index reacts to both organicpollution and habitat diversity changes (Alba-Tercerdorand Sanchez-Ortega, 1988).

Relationships between RHS parameters and biologicalmetrics were tested using correlation analyses. Pearsoncorrelation coefficients were calculated for quantitativemetrics, whereas Spearman rank correlations for a categor-ical metric (HMS class). All correlation analyses were per-formed using SPSS 12.0.

4. Results

4.1. Hydromorphological analyses

The HMS points are increasing linearly in the flow direc-tion from site GLIN 1 to the site GLIN 5 (Fig. 2). Themajority of points was given for the reinforcement of thebanks, only sites GLIN 2 and GLIN 4 have additionalpoints for the presence of an intermediate bridge. Site

Table 2Hydromorphological channel characteristics of the selected sampling sites of the Glinscica stream

Site code Predominant flow type Predominant substrate Current velocity (m/s) Mean width (m) Mean depth (cm)

GLIN 1 Rippled Gravel 0.15 2 14GLIN 2 Smooth + Rippled Silt 0.21 3 28GLIN 3 Smooth Cobble 0.05 3 64GLIN 4 Smooth + Rippled Cobble + Concrete 0.34 2 31GLIN 5 Rippled Concrete 0.52 1 11

0

5

10

15

20

25

30

35

40

45

GLIN 1 GLIN 2 GLIN 3 GLIN 4 GLIN 5

Sampling site

Val

ue

HQAHMS

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

GLIN 1 GLIN 2 GLIN 3 GLIN 4 GLIN 5

Sampling site

Val

ue

HMS classSI

Fig. 2. Results of the River habitat survey and Saprobic index values of the selected sampling sites of the Glinscica stream.

1064 B. Tavzes et al. / Physics and Chemistry of the Earth 31 (2006) 1062–1074

Autho

r's

pers

onal

co

py

GLIN 1 has 0 points for habitat modification and is nearlypristine, whereas GLIN 5 has 40 points and is categorisedas significantly modified. In addition, classification of sitesin HMS classes showed that both sites GLIN 5 and GLIN4 belong to the worst class, class 4. Sites GLIN 2 andGLIN 3 were classified in the class 3, whereas site GLIN1 in the class 1. On the other hand, HQA points for habitatquality are in concordance to the results of the HMS, butless to the HMS classification. The highest numbers ofHQA points were recorded at sites GLIN 1 (38) and GLIN2 (40), the second two sites were given only 19 points eachand at the last site GLIN 5 the lowest number of HQApoints were given, only 6.

4.2. Biological analyses

A total of 57,824 individuals representing 103 macroin-vertebrate taxa were collected. The number of individualsincreased downstream, but at sites GLIN 3 and GLIN 4similar numbers of individuals were recorded. Taxa com-position at each site is given in Fig. 3 and Appendix A.

Gammarus fossarum was the most abundant species, rep-resenting more than half of all individuals at site GLIN 1,where 59 taxa were recorded. Synurella ambulans, Centrop-

tilum luteolum, Tanypodinae and Tanytarsini followed,respectively, whereas none of all the other taxa representedmore than 2.8%. At site GLIN 2 the lowest number of taxa

Fig. 3. Taxonomic composition of macroinvertebrate assemblages at selected sampling sites of the Glinscica stream. (Included are only taxa thataccounted for more than 3% of all individuals at the selected sampling site.)

B. Tavzes et al. / Physics and Chemistry of the Earth 31 (2006) 1062–1074 1065

Autho

r's

pers

onal

co

py

(52) was recorded. G. fossarum was again by far the mostabundant with 72.5%, whereas the second most abundantChironomini represented only 4.6% and any other taxa lessthan 1.9%. The highest number of taxa was recorded at siteGLIN 3 where out of 64 taxa Tanytarsini were the mostabundant and represented 21.5% of all individuals. Eachof the remaining taxa represented less than 3% of individ-uals but together they accounted for 48%. At site GLIN4 similar number of taxa as at site GLIN 3 were recorded.Of 63 taxa, C. luctuosa was the most abundant with 35.5%.There were eight other taxa that accounted for more than3% of individuals. At site GLIN 5 53 taxa were recorded,however C. luctuosa and G. fossarum dominated with46.9% and 41.1%, respectively. Each of all other taxa rep-resented less than 2.5% of the abundance.

Cluster analysis and DCA on the log abundances ofmacroinvertebrate taxa showed that sites GLIN 3 andGLIN 4 are the most similar (Figs. 4 and 5), and that themost pristine site GLIN 1 has the most unique macroinver-tebrate fauna. According to the results of the cluster anal-ysis site GLIN 5 is more similar to sites GLIN 3 and GLIN4 than to sites GLIN 1 or GLIN 2. However, on theF1 · F2 DCA ordination diagram site GLIN 5 is locatedbetween sites GLIN 3 and GLIN 1 and is not furthermostfrom the site GLIN 1.

Longitudinal changes in macroinvertebrate compositionare presented in Fig. 6. At site GLIN 2 an increase in pro-portion of amphipods and oligochaets comparing to siteGLIN 1 was observed. However, the changes are smallerthan those observed at site GLIN 3. At this site only fewamphipods were found, but chironomids dominated.Moreover, an increase in the number of ephemeropteraindividuals was observed, and this trend continued at sitesGLIN 4 and GLIN 5. The same trend was observed alsofor amphipods, but the proportion of chironomids wasdecreasing. For all the other groups there were only minorchanges comparing to the most abundant groups.

Besides the differences in taxa composition samples alsodiffered in the composition of functional feeding groups(Fig. 7). At site GLIN 1, shredders were the most abundantfollowed by detritivores, grazers and predators. At siteGLIN 2, there was a slight increase in the proportion ofdetritivores, but the proportions of all other feeding groupsslightly decreased. At site GLIN 3 almost no shredderswere recorded, but the highest proportions of detritivores,grazers and filterers were observed. At the downstreamsites GLIN 4 and GLIN 5 an increasing trend in the pro-portions of shredders and detritivores was observed, butproportions of all other feeding groups were decreasing.

Also some other calculated macroinvertebrate metricsshowed that differences between selected sites exist. Shan-non–Wiener diversity value (H 0) decreased at site GLIN 2comparing to site GLIN 1, but increased at site GLIN 3 tothe highest value (3.35) observed (Fig. 8). At site GLIN 4diversity did not change much in comparison to site GLIN3, but at site GLIN 5 a large decrease was observed. Thediversity changed from 3.30 (site GLIN 4) to 1.83 (site GLIN

5). IBMWP scores responded differently compared to thediversity index. At site GLIN 1 the highest score wasobtained, whereas at all other sites scores were quite similar.Number of EPT taxa obtained similar results. The highestvalue was observed at site GLIN 1, whereas at sites GLIN2–5 much lower values were observed. However, a slightincrease was observed at the last site. Proportions of EPTtaxa on the other hand gave different results. At site GLIN1, the proportion was 13%. At site GLIN 2, the lowest valuewas obtained, but downstream the proportion of EPT taxaincreased and reached more than 50% at site GLIN 5.

GLIN 3

GLIN 4

GLIN 5

GLIN 2

GLIN 1

00.20.40.60.81

Fig. 4. Clustering results of the macroinvertebrate taxonomic composi-tion using the Bray–Curtis matrix.

GLIN 5

GLIN 4 GLIN 3

GLIN 2

GLIN 1 0

0.2

0.4

0.6

0.8

1

1.2

1.4

0 0.5 1 1.5 2 2.5

DCA 1

DC

A 2

Fig. 5. DCA ordination diagram with the selected sampling sites of theGlinscica stream. Arrows indicate a direction of habitat degradationgradient.

1066 B. Tavzes et al. / Physics and Chemistry of the Earth 31 (2006) 1062–1074

Autho

r's

pers

onal

co

pyThe saprobic index shows that the organic pollution of

the Glinscica stream is moderate and belongs to the II.quality class, although, at site GLIN 1 the lowest valuewas calculated (Fig. 2). At site GLIN 5, a slight decreasein organic pollution comparing to site GLIN 4 wasobserved.

4.3. Relationship between RHS parameters and biological

metrics

All selected RHS parameters had at least two statisti-cally significant correlations with selected biological met-rics (Table 3). However, differences were found in thenumber of metrics and in the parameters that correlated.The HMS had statistically significant correlation with fourmetrics, whereas HQA and HMS class with three and two,respectively (Fig 9). The highest correlation (r = 0.99,P < 0.01) was between HMS and % of detritivores. HMSwas also correlated with log number of individuals(r = 0.98, P < 0.01), % of C. luctuosa (r = 0.95, P < 0.05)and % of EPT individuals (r = 0.89, P < 0.05). On theother hand, HMS class was correlated only with log num-

ber of individuals (r = 0.95, P < 0.05) and % of detritivores(r = 0.94, P < 0.05). HQA was similarly as HMS correlatedwith % of detritivores (r = 0.93, P < 0.05), % of C. luctuosa

(r = 0.91, P < 0.05) and % of EPT taxa (r = 0.92, P < 0.05),but values of correlation coefficients were slightly different.

5. Discussion

Urbanized lotic systems are often classified as havingpoor or very poor biotic integrity (Davis et al., 2003).Organic pollution is a well known factor that affect macro-invertebrate assemblages in rural as well as in urban rivers(Bascombe et al., 1989; Beavan et al., 2001). According tothe calculated saprobic index the Glinscica stream is mod-erately polluted at all selected sampling sites. Therefore, inour study organic pollution was excluded as being an influ-ential factor causing differences in macroinvertebrateassemblages. Physical habitat changes are the other fac-tor that significantly affects urban rivers (Beavan et al.,2001). Differences in physical habitat between selected sam-pling sites on the Glinscica stream are evident. At siteGLIN 1 the stream is nearly pristine with natural riparian

0%

20%

40%

60%

80%

100%

GLIN 1 GLIN 2 GLIN 3 GLIN 4 GLIN 5

Sampling site

Tax

on

DIPTERA

TRICHOPTERA

COLEOPTERA

ODONATA

PLECOPTERA

EPHEMEROPTERA

OTHER INSECTS

CRUSTACEA

HIRUDINEA

OLIGOCHAETA

MOLLUSCA

PLATHELMINTHES

Fig. 6. Composition of the macroinvertebrate assemblages at selected sampling sites of the Glinscica stream.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

GLIN 1 GLIN 2 GLIN 3 GLIN 4 GLIN 5

Sampling site

Fu

nct

ion

al f

eed

ing

gro

up

OTHER

PARASITS

PREDATORS

DETRITIVORES

FILTERERS

GRAZERS

SHREDDERS

Fig. 7. Composition of the functional feeding groups at selected sampling sites of the Glinscica stream.

B. Tavzes et al. / Physics and Chemistry of the Earth 31 (2006) 1062–1074 1067

Autho

r's

pers

onal

co

py

vegetation, whereas downstream sites are located at alteredsites from which natural vegetation has been removed(Table 1). Riparian vegetation is an important feature in

the structure and function of lotic systems, providingcover, shading, nutrient inputs, woody debris, terrestrialhabitat, and buffering against contaminant inputs fromrunoff (Lazorchak et al., 1998).

Results of the RHS confirm a general impression thatthe stream, as a habitat, is deteriorating from site GLIN1 to GLIN 5 (Fig. 2). Reinforcement of the banks andchannelisation, which progresses from site GLIN 2 to siteGLIN 5, contributes to the rise in HMS points. Addition-ally the decline in HQA points confirmed that the Glinscicastream is deteriorating as a stream habitat because of thechanges in the channel as well as the anthropogenic land-use disturbances in the associated habitats. The mostupstream site scored the highest number of points for treesand associated features. As trees and riparian vegetationbecomes scarcer in the flow direction, so does the amountof shading and woody debris and consequently the numberof HQA points declines. When the overhead canopy of astream is lost, both the shade that controls temperatureand the supply of leaf litter entering the aquatic food chainis eliminated (Booth and Jackson, 1997). In places wherethe amount of shading is lower the amount of channel veg-etation increases. At the site GLIN 5 there is no shading,

1

10

100

1000

10000

100000

GLIN 1 GLIN 2 GLIN 3 GLIN 4 GLIN 5

Sampling site

Ab

un

dan

ce

0

50

100

150

200

250

GLIN 1 GLIN 2 GLIN 3 GLIN 4 GLIN 5

Sampling site

IBM

WP

0

10

20

30

40

50

60

70

GLIN 1 GLIN 2 GLIN 3 GLIN 4 GLIN 5

No

. tax

a

0

2

4

6

8

10

12

14

16

18

GLIN 1 GLIN 2 GLIN 3 GLIN 4 GLIN 5

Sampling site

No

. EP

T t

axa

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

4.00

GLIN 1 GLIN 2 GLIN 3 GLIN 4 GLIN 5

H'

0

10

20

30

40

50

60

GLIN 1 GLIN 2 GLIN 3 GLIN 4 GLIN 5

Sampling site

% E

PT

ind

ivid

ual

s

a d

b e

c f

Sampling site

Sampling site

Fig. 8. Macroinvertebrate metrics at selected sampling sites of the Glinscica stream. (a) Macroinvertebrate abundance (log scale), (b) number of taxa,(c) diversity values (Shannon–Wiener), (d) IBMWP scores, (e) number of EPT taxa and (f) percentage of EPT individuals.

Table 3Spearman rank and Pearson correlation coefficients (r) between RHSparameters and biological metrics and their statistical significance (P)

HQA HMS HMS class

No. of individuals (log) �0.80 0.98** 0.95*

No. of taxa �0.15 �0.13 0.00No. of families �0.06 �0.28 �0.46H 0 �0.10 �0.17 0.00IBMWP 0.30 �0.53 �0.32% EPT individuals �0.92* 0.89* 0.79No. of EPT taxa 0.19 �0.33 �0.11% Caenis luctuosa �0.91* 0.95* 0.79% Shredders 0.72 �0.50 �0.53% Grazers �0.25 �0.05 �0.21% Filterers �0.37 0.08 0.00% Detritivores �0.93* 0.99** 0.95*

% Predators 0.48 �0.77 �0.74% Parasites �0.19 �0.09 0.18% Other �0.45 0.17 0.53

* Correlation is significant at the 0.05 level (two-tailed).** Correlation is significant at the 0.01 level (two-tailed).

1068 B. Tavzes et al. / Physics and Chemistry of the Earth 31 (2006) 1062–1074

Autho

r's

pers

onal

co

py

but because of the concrete reinforcement of the channelthe instream vegetation is limited to filamentous algaeand mosses. Only one flow type is present at this site,whereas in the other sites at least three flow types are pres-ent. The same goes for channel substrate.

Along some stretches of the Glinscica stream, themacroinvertebrate community did not follow this obviouslongitudinal pattern in habitat deterioration. Results ofsimilarity analyses of taxonomic composition (Figs. 4–6)as well as comparison of functional feeding groups compo-sition (Fig. 7) showed that macroinvertebrate assemblagesfollowed the longitudinal habitat degradation pattern onlyat the first three sites, but not at the last two sites (Fig. 5).At site GLIN 4 a concrete substrate was partially presentand the water depth was lower than at the previous sam-pling site, which resulted in the flow type change (Table2). Moreover, at site GLIN 5 only concrete substrate was

present and the water was very shallow. These changes inhydromorphology were probably the reason for macroin-vertebrate assemblages at the last two sites being more sim-ilar to the upstream sites than they should have, accordingto the habitat quality results assessed with RHS. Pedersenand Perkins (1986) did not find any differences in macroin-vertebrate abundances between urban and rural streams. Inour study, the number of individuals increased downstreamand at the most physical altered site the highest numberof individuals was found. However, only two speciesdominated: C. luctuosa and G. fossarum. These are speciesthat are tolerant to hydrologically unstable urban streamenvironments (Gonzalez et al., 2001; Peran et al., 1999;Pockl et al., 2003) and in the regulated parts of the Glin-scica stream, were associated mainly with mosses alongthe channel edge. On the other hand, at the undisturbedsite GLIN 1 many different taxa were recorded and many

r = 0.98P<0.01

0.00 10.00 20.00 30.00 40.00

HMS

3.4

3.6

3.8

4

4.2

4.4

No

. in

div

idu

als

(lo

g)

r = 0.99 P<0.01

0.00 10.00 20.00 30.00 40.00

HMS

0.20

0.30

0.40

0.50

0.60

% D

etri

tivo

res

r = -0.92 P<0.05

0.00 10.00 20.00 30.00 40.00

HQA

0

0.1

0.2

0.3

0.4

0.5

% E

PT

ind

ivid

ual

s

r = 0.89 P<0.05

0.00 10.00 20.00 30.00 40.00

HMS

0

0.1

0.2

0.3

0.4

0.5

% E

PT

in

div

idu

als

r = 0.95 P<0.05

1.00 1.50 2.00 2.50 3.00 3.50 4.00

HMS class

3.400

3.600

3.800

4.000

4.200

4.400

No

. in

div

idu

als

(lo

g)

r = -0.91 P<0.05

0.00 10.00 20.00 30.00 40.00

HQA

0.00

10.00

20.00

30.00

40.00

50.00

% C

aen

is lu

ctu

osa

r = 0.95 P<0.05

0.00 10.00 20.00 30.00 40.00

HMS

0.00

10.00

20.00

30.00

40.00

50.00

% C

aen

is lu

ctu

osa

r = 0.95 P<0.05

1.00 1.50 2.00 2.50 3.00 3.50 4.00

HMS class

0.20

0.30

0.40

0.50

0.60

% D

etri

tivo

res

r = -0.93 P<0.05

0.00 10.00 20.00 30.00 40.00

HQA

0.20

0.30

0.40

0.50

0.60

% D

etri

tivo

res

a d g

b e h

c f i

Fig. 9. Scatter plots of statistically significant correlations between RHS parameters and macroinvertebrate metric.

B. Tavzes et al. / Physics and Chemistry of the Earth 31 (2006) 1062–1074 1069

Autho

r's

pers

onal

co

py

of them were recorded only there. These were mainly EPTtaxa, which are sensitive to habitat modification. The per-centage of EPT individuals is commonly used and recom-mended for a measure of stream quality (Quinn et al.,1997; Davis et al., 2003; Sandin and Johnson, 2000;AQEM, 2002). Moreover, they (op. cit.) observed thatthe number of EPT individuals decreases in poorer condi-tions. However, we observed a negative correlationbetween percentage of EPT individuals and HQA pointsand positive correlation between percentage of EPT indi-viduals and HMS points. Percentage of C. luctuosa namelyincreased drastically from site GLIN 3 to GLIN 5 wherethey accounted for 47% of all individuals (Fig. 3). Obvi-ously C. luctuosa does not respond to habitat degradationas do the majority of EPT taxa. Gonzalez et al. (2001)found that C. luctuosa had asynchronic life cycles withextended recruitment periods, which increases the proba-bility for sustaining vital populations in unstable flowregimes that is characteristic for modified parts of the Glin-scica stream. Correlations between the abundance of C.luctuosa and HQA or HMS show that relative abundanceof this species could be used as a measure of habitatdegradation.

In degraded parts of the stream we anticipated thatchironomids and oligocheats would rise in their number.But at sites GLIN 4 and GLIN 5 there was a rise inephemeroptera and amphipods. It is evident that these sitesare morphologically modified and thus uniform in habitatbut the level of organic pollution is moderate. Those cir-cumstances obviously enable proliferation of the presenttaxa. G. fossarum is known to inhabit small shallowstreams with high current velocities with low concentra-tions of nutrients (Peeters and Gardeniers, 1998; Pocklet al., 2003). G. fossarum has been quite abundant in GLIN1 and 2 as well as at site GLIN 5. We presume that they usedifferent food resources for their diet. Gammarus are possi-ble to change their diets and are omnivorous, utilising awide variety of foodstuffs to maximise fitness (Kellyet al., 2002). From the composition of functional feedinggroups we can see that the percentage of detritivoresfollows the pattern of habitat changes, which increasesfrom site GLIN 1 to GLIN 5. The percentage of shreddersdecline drastically at site GLIN 3 where a decline of ripar-ian vegetation is also present. On the other hand, the

percentage of grazers rises at that point as a result ofemerging instream vegetation.

According to IBMWP scores, site GLIN 1 has the high-est ecological value, whereas IBMWP scores of all regulatedsites were much lower. That less modified sites support aslightly higher quality of macroinvertebrate fauna was alsoobserved by Beavan et al. (2001). They confirmed this withhigher calculated BMWP scores and beta diversity, but notwith alpha diversity, which was very similar in regulatedand unregulated sites. However, at regulated sites GLIN 3and GLIN 4 the highest number of taxa and Shannon–Wie-ner diversity were recorded. That slightly altered plane riv-ers with stones and riffles support high number of taxa anddiversity was observed also for Trichoptera in the ScavnicaRiver (Urbanic, 1999; Urbanic et al., 2000, 2005). However,mainly very common species were found, which means thatdespite higher diversity and taxa richness the quality of thefauna was lower. Similar situations have been observed alsoin the Glinscica stream. Taxa found at sites GLIN 3 andGLIN 4 are very common.

6. Conclusions

1. Macroinvertebrate taxonomic composition and func-tional feeding groups composition in the Glinscicastream did not follow a degradation pattern in thestream habitat. Pristine and the most altered sites werenot the most diverse.

2. Stable substrate (concrete) and shallow rippled water flowcan partially reduce the influence of the habitat degrada-tion on macroinvertebrate community, which resulted inhigh similarity of the most degraded and pristine sites.

3. Percentage of detritivores, percentage of C. luctuosa,number of individuals, percentage of EPT individualsare best predictors of changes in the physical habitatquality. However, the metric percentage of EPT individ-uals was negatively correlated to the habitat degrada-tion, which is in contradiction with results from studiesof other authors.

Appendix A

See Table A.1.

Table A.1Taxonomic composition and relative abundance (%) of macroinvertebrates at selected sampling sites of the Glinscica stream

Taxa Sampling site

GLIN 1 GLIN 2 GLIN 3 GLIN 4 GLIN 5

Relative abundance (%)

PLATHELMINTHESTurbellariaFam. Planariidae

Polycelis tenuis/nigra 0.5 0.2Fam. Dugesiidae

Dugesia sp. 0.1 0.4 0.1

1070 B. Tavzes et al. / Physics and Chemistry of the Earth 31 (2006) 1062–1074

Autho

r's

pers

onal

co

py

Table A.1 (continued)

Taxa Sampling site

GLIN 1 GLIN 2 GLIN 3 GLIN 4 GLIN 5

MOLLUSCAGastropodaFam. Hydrobiidae

Sadleriana fluminensis (Kuester, 1853) 0.8Fam. Lymneaidae

Galba truncatula (O.F. Muller, 1774) 0.7 0.7Radix balthica (Linnaeus, 1758) 1 0.5Radix labiata (Rossmassler, 1835) 0.5 0.6

Fam. PhysidaePhysa fontinalis (Linnaeus, 1758) 0.3 0.1

Fam. PlanorbidaeGyraulus albus (O.F. Muller, 1774) 2.4 2.7 0.1Gyraulus crista (Linnaeus, 1758) 1 0.5Hippeutis complanatus (Linnaeus, 1758) 0.9 0.4 0.4Planorbis planorbis (Linnaeus, 1758) 0.1 0.4 0.3

Fam. AncylidaeAncylus fluviatilis (O.F. Muller, 1774) 0.2 0.5

BivalviaFam. Sphaeridae

Pisidium sp. 1.3 0.4 0.3 0.2 0.4

CLITELLATAOligochaetaFam. Tubificidae

Tubificidae – with hair chaeta 0.6 1.5 0.4 0.7Tubificidae – without hair chaeta 0.3 0.3 0.7 0.7

Fam. LumbriculidaeLumbriculus variegatus (Muller, 1774) 0.8 0.4 0.1Stylodrilus heringianus (Claperede, 1862) 1.8 1.2 2.2 0.4

Fam. LumbricidaeEiseniella tetraedra (Savigny, 1826) 0.1 0.8 0.2 0.3 0.3

HirudineaFam. Glossiphonidae

Glossiphonia complanata (Linnaeus, 1758) 0.1 0.1 0.1Helobdella stagnalis (Linnaeus, 1758) 0.1 0.8 0.8 0.4Hemiclepsis marginata (O.F. Muller, 1774) 0.7

Fam. ErpobdellidaeErpobdella octoculata (Linnaeus, 1758) 0.7 0.6 0.4 0.8Erpobdella testacea (Savigny, 1822) 0.4Theromyzon tessulatum (O.F. Muller, 1774) 0.1

CRUSTACEAAmphipodaFam. Crangonyctidae

Synurella ambulans (Fr. Muller, 1846) 13.4Fam. Gammaridae

Gammarus fossarum (Koch in Panzer, 1836) 53.5 72.5 0.3 7.9 41.1IsopodaFam. Asellidae

Asellus aquaticus (Linnaeus, 1758) 0.8 0.3 0.2DecapodaFam. Astacidae

Austropotamobius torrentium (Schrank, 1803) 0.3 0.5 0.8

INSECTAEphemeropteraFam. Baetidae

Baetis liebenauae (Keffermuller, 1972) 0.1Baetis rhodani (Pictet, 1843–1845) 0.3 0.2 1 0.2Baetis fuscatus/scambus 0.7Baetis vernus (Curtis, 1834) 0.9 6.1 3.2 2.3Centroptilum luteolum (Muller, 1776) 5.5Cloeon dipterum (Linnaeus, 1761) 1.6 0.6

(continued on next page)

B. Tavzes et al. / Physics and Chemistry of the Earth 31 (2006) 1062–1074 1071

Autho

r's

pers

onal

co

py

Table A.1 (continued)

Taxa Sampling site

GLIN 1 GLIN 2 GLIN 3 GLIN 4 GLIN 5

Fam. HeptagenidaeEcdyonurus sp. 2.1

Fam. EphemerellidaeEphemerella ignita (Poda, 1761) 0.3

Fam. CaenidaeCaenis horaria (Linnaeus, 1758) 0.2 0.3 0.4Caenis luctuosa (Burmeister, 1839) 0.3 12.5 35.5 46.9

Fam. LeptohlebiidaeParaleptohlebia submarginata (Stephens, 1835) 0.6 0.4Habroleptoides confusa (Sartori and Jacob, 1986) 1.3Habrophlebia lauta (Eaton, 1884) 0.4

Fam. EphemeridaeEphemera danica (Muller, 1764) 0.3 0.8

PlecopteraFam. Perlodidae

Isoperla sp. 0.3Fam. Nemouridae

Nemoura sp. 0.2Amphinemura sp. 0.7

Fam. LeuctridaeLeuctra sp. 0.8

Fam. PerlidaePerla sp. 0.3

OdonataFam. Calopterygidae

Calopteryx virgo (Linnaeus, 1758) 0.5 0.3 0.1Fam. Platycnemididae

Platycnemis pennipes (Pallas, 1771) 0.2 1.4 0.1Fam. Coenagrionidae 0.2 0.6

Coenagrion sp. 0.1 0.7Ischnura sp. 0.3

Fam. GomphidaeOnychogomphus forcipatus (Linnaeus, 1758) 0.3 0.4 0.1

Fam. CorduliidaeSomatochlora meridionalis (Nielsen, 1935) 0.6 0.8

Fam. LibellulidaeOrthethrum coerulescens (Fabricius, 1798) 0.7

ColeopteraFam. Dytiscidae 0.3 0.4

Agabus sp. 0.3Platambus maculatus (Linnaeus, 1758) 0.3 0.5 0.3 0.3

Fam. Hydrophilidae 1 0.5Fam. Hydraenidae

Hydraena sp. 0.4 0.2 1 1 0.5Fam. Dryopidae

Dryops sp. 0.1Fam. Elmidae

Limnius sp. 0.1 0.2 0.1 0.3 0.4Esolus sp. 0.7 0.1 0.1 0.4Elmis sp. 0.7 0.7 1 1 0.2Oulimnius sp. 1.2 1.3 3.8 2.4

Fam. HaliplidaeHaliplus sp. 0.6 7.8 3.7 1

Fam. Curculionidae 1 0.4HeteropteraFam. Notonectidae

Notonecta sp. 0.1Fam. Nepidae

Nepa sp. 0.7 0.4Fam. CorixidaeCorixinae 0.1MegalopteraFam. Sialidae

Sialis fuliginosa (F.J. Pictet, 1836) 0.7 0.2

1072 B. Tavzes et al. / Physics and Chemistry of the Earth 31 (2006) 1062–1074

Autho

r's

pers

onal

co

py

References

Alba-Tercerdor, J., Sanchez-Ortega, A., 1988. Un metodo rapido y simplepara evaluarla calidad biologica de las aguas corrientes basado en el deHelawell (1978). Limnetica 4, 51–56.

AQEM Consortium, 2002. Manual for the application of the AQEMsystem. A comprehensive method to assess European streams usingbenthic macroinvertebrates, developed for the purpose of the Water

Framework Directive. Version 1.0. (February, 2002). Available from:<http://www.aqem.de/start.htm> (March, 2002).

Bascombe, A.D., Ellis, J.B., Revitt, D.M., Shutes, R.B.E., 1989. Macr-oinvertebrate biomonitoring and water quality management withinurban catchments. Technical report, Middlesex Polytechnic, London,pp. 403–411.

Beavan, L., Sadler, J., Pinder, L., 2001. The invertebrate fauna of aphysically modified urban river. Hydrobiologia 445, 97–108.

Table A.1 (continued)

Taxa Sampling site

GLIN 1 GLIN 2 GLIN 3 GLIN 4 GLIN 5

PlanipenniaFam. Sysiridae

Sysira sp. 0.8TrichopteraFam. Rhyacophilidae

Rhyacophila fasciata (Hagen, 1859) 0.4Fam. Hydroptilidae

Hydroptila sparsa (Curtis, 1834) 0.2 0.2 0.2Hydroptila sp. 0.4

Fam. HydropsychidaeHydropsyche angustipennis (Curtis, 1834) 0.5

Fam. PolycentropodidaePlectrocnemia conspersa (Curtis, 1834) 1.8

Fam. PsychomyiidaeTinodes unicolor (Pictet, 1834) 0.3

Fam. LimnephilidaeMicropterna nycterobia (McLachlan, 1875) 0.7

Fam. LeptoceridaeMystacides azurea (Linnaeus, 1761) 0.7 0.2 0.6 0.4

Fam. SericostomatidaeNotidobia ciliaris (Linnaeus, 1761) 0.4

DipteraFam. Psychodidae 0.1 0.5Fam. Simuliidae

Simulium sp. 0.3 0.9 1.4 3.5 1.5Fam. CeratopogonidaeCeratopogoninae 0.7 0.2 0.5 0.4 0.5Forcipomyinae 0.1Fam. ChironomidaeTanypodinae 5.3 1.9 2.7 3.2 1ChironominaeChironomini 0.3 4.6 1.8 0.3 0.4Tanytarsini 4.6 1 21.5 13.9 1.4Orthocladiinae 2.8 0.5 3.6 16.1 1.7Diamesinae 0.3Prodiamesinae 0.2

Prodiamesa olivacea (Meigen, 1818) 0.3 0.2Fam. Tipulidae

Tipula sp. 1 0.2 0.1 0.6Fam. Limoniidae 0.2Hexatominae 0.4 0.1Fam. Pediciidae

Dicranota sp.Pedicia sp. 0.1

Fam. EmpididaeClinocerinae 0.3Fam. Stratiomyidae

Beris sp. 0.3Oxycera sp. 0.1

Fam. Tabanidae 0.7 0.1 0.1 0.4Fam. Muscidae

Limnophora sp. 0.9 0.4 0.5

B. Tavzes et al. / Physics and Chemistry of the Earth 31 (2006) 1062–1074 1073

Autho

r's

pers

onal

co

py

Beisel, J.-N., Usseglio-Polatera, P., Moreteau, J.C., 2000. The spatialheterogeneity of a river bottom: a key factor determining macroinver-tebrate communities. Hydrobiologia 422–423, 163–171.

Booth, D.B., Jackson, C.R., 1997. Urbanization of aquatic systems:degradation thresholds, stormwater detection, and the limits of mitiga-tion. Journal of the American Water Resources Association 22, 1–20.

Brookes, A., Gregory, K.J., 1998. Channelisation, river engineering andgeomorphology. In: Hooke, J.M. (Ed.), Geomorphology in Environ-mental Planning. J. Wiley & Sons, Chichester, pp. 68–145.

Buffagni, A., Crosa, G.A., Harper, D.M., Kemp, J.L., 2000. Usingmacroinvertebrate species assemblages to identify river channel habitatunits: an application of the functional habitats concept to a large,unpolluted Italian river (River Ticino, northern Italy). Hydrobiologia435, 213–225.

Davis, N.M., Weaver, V., Parks, K., Lydy, M.J., 2003. An assessment ofwater quality, physical habitat, and biological integrity of an urbanstream in Wichita, Kansas, prior to restoration improvements (Phase I).Archives of Environmental Contamination and Toxicology 44, 351–359.

Environment Agency, 2002. River Habitat Survey Database, Version 3.3.Environment agency, 2003. River Habitat Survey in Britain and Ireland,

Field Survey Guidance Manual, 2003 Version.EU, 2000. Directive 2000/60/EC of the European Parliament and of the

Council of 23 October 2000 establishing a framework for Communityaction in the field of water policy. Official Journal of the EuropeanCommunities L 327, 22 December 2000, pp. 1–72.

Gonzalez, J.M., Basaguren, A., Pozo, J., 2001. Life history and produc-tion of Caenis luctuosa (Burmeister) (Ephemeroptera, Caenidae) in twonearby reaches along a small stream. Hydrobiologia 452, 209–215.

Kelly, D.W., Dick, J.T.A., Montgomery, W.I., 2002. The functional roleof Gammarus (Crustacea, Amphipoda): shredders, predators, or both?Hydrobiologia 485, 199–203.

Lazorchak, J.M., Klemm, D.J., Peck, D.V., 1998. Environmental Mon-itoring and Assessment Program—Surface Waters: Field Operationsand Methods for Measuring the Ecological Condition of WadeableStreams. EPA/620/R-94/004F, US Environmental Protection Agency,Office of Research and Development, Washington, DC.

Lorenz, A., Hering, D., Feld, C.K., Rolauffs, P., 2004. A new method forassessing the impact of hydromorphological degradation on themacroinvertebrate fauna of five German stream types. Hydrobiologia516, 107–127.

Moog, O. (Ed.), 1995. Fauna Aquatica Austriaca, Version 1995. Wasser-wirtschafts-kataster, Bundesministerium fur Land-und Forstwirts-chaft, Wien.

Pedersen, E.R., Perkins, M.A., 1986. The use of benthic invertebrate datafor evaluating impacts of urban runoff. Hydrobiologia 139, 13–22.

Peran, A., Velasco, J., Millan, A., 1999. Life cycle and secondaryproduction of Caenis luctuosa (Ephemeroptera) in a semiarid stream(Southeast Spain). Hydrobiologia 400, 187–194.

Peeters, E.T.H.M., Gardeniers, J.J.P., 1998. Logistic regression as a toolfor defining habitat requirements of two common gammarids. Fresh-water Biology 39, 605–615.

Pockl, M., Webb, B.W., Sutcliffe, D.W., 2003. Life history and reproduc-tive capacity of Gammarus fossarum and G. roeseli (Crustacea:Amphipoda) under naturally fluctuating water temperatures: a simu-lation study. Freshwater Biology 48, 53–66.

Quinn, J.M., Cooper, A.B., Davies-Colley, R.J., Rutherford, J.C.,Williamson, R.B., 1997. Land use effects on habitat, water quality,periphyton, and benthic invertebrates in Waikato, New Zealand hillcountry streams. New Zealand Journal of Marine and FreshwaterResearch 31, 579–597.

Rusjan, S., Fazarinc, R., Mikos, M., 2003. River rehabilitation of urbanwatercourses on the example of the glinscica river in ljubljana. ActaHydrotechnica (21/34), 1–22.

Sandin, L., Johnson, R.K., 2000. The statistical power of selectedindicator metrics using macroinvertebrates for assessing acidificationand eutrophication of running waters. Hydrobiologia 422–423, 233–243.

Urbanic, G., 1999. Taxonomical and ecological characteristics of Tri-choptera community in the River Scavnica. Graduation Thesis,University of Ljubljana, Biotechnical Faculty, 69 pp.

Urbanic, G., 2004. Ecology and distribution of caddisflies (Insecta:Trichoptera) in some watercourses in Slovenia. Dissertation Thesis,University of Ljubljana, Biotechnical Faculty, 189 pp.

Urbanic, G., Toman, M.J., Krusnik, C., 2000. Downstream changes incaddisfly fauna (Insecta, Trichoptera) in relation to environmentalvariables in the River Scavnica (NE Slovenia): a multivariate approachusing a Canonical Correspondence Analysis (CCA). Acta BiologicaSlovenica 43 (4), 21–35.

Urbanic, G., Toman, M.J., Krusnik, C., 2005. Microhabitat type selectionof caddisfly larvae (Insecta: Trichoptera) in a shallow lowland stream.Hydrobiologia 541, 1–12.

Urbanic, G., 2005. Abiotska tipizacija vodotokov. V: Urbanic G. (ur.)Ekolosko stanje za reke in jezera, porocilo o delu v letu 2005. Institutza vode Republike Slovenije, Ljubljana, 193 pp. (in Slovenian).

Wegl, R., 1983. Index fur die Limnosaprobitat. Wasser und Abwasser 26,1–175.

Zelinka, M., Marvan, P., 1961. Zur Prazisierung der biologischenKlassifikation der Reinheit fließender Gewasser. Archiv fur Hydrobi-ologie 57, 389–407.

1074 B. Tavzes et al. / Physics and Chemistry of the Earth 31 (2006) 1062–1074