Ecology and diet of the endemic ukliva () in the Cetina River system, Croatia

by

Davor Zanella*, Zlatko MihaljeviĆ*, Milorad MrakovČiĆ & Marko Ćaleta (1)

ABSTRACT. - The first data on the population ecology and diet of the endemic ukliva dace (Telestes ukliva Heckel, 1843) are presented. Along the Cetina River system, the ukliva dace accounted for 38.4% of the overall fish fauna abundance, but only 6.4% of total fish biomass.T he ukliva dace population exhibited migratory behaviour during the year and showed a strong temporal component; inhabiting the river main course in spring and summer, and colonising springs and subterrane- an waters in late autumn, where most of population overwintered. Males are smaller than females, while the sex ratio at the population level was 1:1.6 in favour of males, though this varied by station. The diet can be divided into two phases; the first in the warmer part of the year when feeding is intensive and Simuliidae larvae dominated in gut contents.T his was fol- lowed by the winter and early spring period, when feeding markedly decreased due to its predominant distribution in over- wintering habitats (underground waters). The highest frequency of empty stomachs and unidentifiable digested matter and detritus occurred in early spring. Although some potential pray (Coleoptera) were present in high densities of the substra- tum of Cetina River, they were not found in the stomach contents of the ukliva dace.

résumé. - Écologie et alimentation de Telestes ukliva, espèce endémique de la rivière Cetina, Croatie. Les premières données sur l’écologie de la population et le régime alimentaire de l’espèce endémique Telestes ukliva Heckel, 1843 sont présentées. Le long de la rivière Cetina, T. ukliva représente 38,4% de l’ensemble de l’abondance fauni- que en poissons, mais seulement 6,4% de la biomasse totale en poissons. La population a manifesté un comportement migratoire au cours de l’année. La distribution de la population a une composante temporelle marquée; elle habite le cours principal de la rivière au printemps et en été et elle colonise les sources et les eaux souterraines à la fin de l’automne. Les mâles sont plus petits que les femelles, le sex-ratio étant de 1: 1,6 en faveur des mâles, quoique cela ait varié selon la sta- tion. Le régime alimentaire peut être divisé en deux phases. Durant la période estivale, plus chaude, l’alimentation est intensive et les larves de Simuliidae prédominent dans le contenu des intestins. En hiver et au début du printemps, l’alimen- tation a diminué de façon marquée en raison de la distribution prédominante des individus dans les habitats d’hivernage (les eaux souterraines). On note au début du printemps la fréquence la plus élevée d’estomacs vides, de matière digérée non-identifiable et de détritus. Bien que quelques proies potentielles (Coleoptera) soient présentes en hautes densités dans le substratum de la rivière Cetina, aucune n’a été trouvée dans les estomacs de T. ukliva.

Key words. - Cyprinidae - Telestes ukliva - Croatia - Cetina River system - Migratory behaviour - Sex ratio - Diet.

The Dinaric region is recognized as an European biodi- in the Croatian Red Book (Mrakovčić et al., 2006). versity hotspot (Gaston and Davis, 1994) and renowned for Telestes ukliva (Heckel, 1843), the ukliva dace, was pre- its high endemism, especially in aquatic (Bănărescu, 2004) viously included in the genus Leuciscus, but following a and subterranean fauna (Sket et al., 2004). With 151 fresh- rehabilitation of the cyprinid genus Telestes Bonaparte, 1837 water fish species, Croatia is one of the most diverse coun- (Ketmaier et al., 1998), it was reassigned to that genus. tries in Europe. Many of these species are endemic, with 44 T. ukliva, endemic to the Cetina River in southern Croatia species endemic to the Mediterranean, of which 41 species was assessed to be extinct in 2005 (Smith and Darwall, are endemic to the Adriatic basin and are not found else- 2006). However, a recent re-evaluation of the status of the where (Mrakovčić et al., 2006). These endemic species are ukliva dace confirmed that it is indeed present in the Cetina closely tied to their distinctive karst habitats. The Dalmatian River and seasonally abundant (Zanella et al., 2008). Its region is particularly significant, with a high number of IUCN status in Croatia has been classified as critically endemic fish species inhabiting its river basins. One of the endangered (Mrakovčić et al., 2006); however, this is yet to reasons for the large number of endemics is the fact that this be revised, with Kottelat and Freyhof (2007) suggesting a area was not greatly affected by glaciations. Freshwater fish classification of vulnerable (VU D2). are among the most threatened groups of vertebrates and No studies on the biology or ecology of this species exist. approximately 60% of the Adriatic species have been listed A monitoring programme of the Cetina River was conducted

(1) Department of Zoology, Faculty of Science, University of Zagreb, Rooseveltov trg 6, 10000 Zagreb, CROATIA. [[email protected]] * Both should be considered as equivalent authors.

Cybium 2009, 33(2): 97-105. Ecology and diet of Telestes ukliva Za n e l l a e t a l . from 1998-2001 to inventory the fish fauna in the river, and the species T. ukliva was one of the main interests of the monitoring study. The objective of the present study was to determine the relative abundance and biomass of ukliva dace, sex ratio and to establish its seasonal feeding patterns in habitats where the population density is high and the spe- cies is most common.

MATERIALS AND METHODS

Study area and sites characteristics The only known population of ukliva dace is in the Ceti- na River, a karst river of the Adriatic watershed, flowing 105 km from the several karst springs to the river’s mouth at the town of Omiš. The Cetina River is a typical watercourse in the deep and well-developed Dinaric karst, an area of limestone and mainly underground drainage, marked by a characteristic topography. There are strong underground communications with other waters in neighbouring Bosnia and Herzegovina (Crkvenčić et al., 1974). The river system includes five hydroelectric power plants, five reservoirs, and three long tunnels and pipelines. The construction of hydro- electric structures has resulted in changes to the hydrological regime of permanent karst springs on the left tributaries of the river. The presence of hydroelectric plants has greatly stabilized the water regime of the river, though the mean annual discharge has decreased (Bonacci and Roje-Bonacci, 2003). The climate in the upper reaches is continental, while Figure 1. - Map of the study area; stations: S1, S2-Han, S3-Krenica in the lower reaches it is maritime, with some features of a spring, S4-Jezerce spring, S5-Čikotina lađa, S6-Blato na Cetini, Mediterranean climate. The mean annual air temperature is S7-Radmanove mlinice; PP: hydroelectric power plant. between 5 and 8°C in the eastern upper reaches and between 11 to 13°C in the western lower reaches of the river (Bonacci ified by man and are encircled by concrete well heads.T he and Roje-Bonacci, 2003). springs are characterized by relatively constant physical and The study area with sampling sites is shown in figure 1. chemical properties of the water. During the investigation Inventorying during the monitoring study was conducted at period, the water temperature varied from 11°C in winter to seven stations along the entire length of the river, from the 13.2°C in summer, pH ranged from 7.6-7.9 and conductivity springs to the mouth. Ukliva dace is found throughout most from 248-388 µS cm-1. Oxygen saturation was within the of the river, with only a single specimen caught at station S1 range 83-99%. The investigated sites situated at the main (upstream of Peruča reservoir) due to predation by brown river course had a greater variation of physical and chemical trout, the only predatory species in the river source. Telestes properties of the water. Temperature ranged from 4.6°C in ukliva is most abundant in the middle stretch and in the winter to 19.7°C in summer, pH ranged from 7.9-8.4, con- springs on the left tributaries of the Cetina River (Mrakovčić ductivity from 320-484 µS cm-1 and oxygen saturation et al., 2001). The population structure analysis was per- exceeded 100%. The station S1 has a natural hydrological formed at stations S3 (Krenica spring), S4 (Jezerce spring), regime, and the river is approx. 20 m wide, and 0.3-1.0 m S5 (Čikotina lađa) and S6 (Blato na Cetini), while diet anal- deep. The substrate is stony, with the presence of cobbles yses were conducted at S3, S4 and S6. and pebbles. At station S2 (Han), the river is approx. 50 m Krenica spring (station S3) and Jezerce spring (S4) are wide and 2-2.5 m deep with a varied daily flow regime, due small karst springs about 0.5 km apart, that form small to the influence of the Peruča hydroelectric plant.T he sub- watercourses that drain into the Cetina River. Krenica is a strate is pebbles and gravel. Station S5 (Čikotina lađa) is in limnocrenic spring, approx. 12 m wide, 5 m long and 0.3- the canyon section of the river, just downstream from the 0.5 m deep. Jezerce is a linear spring, measuring 2 m wide, Đale hydroelectric plant, and so the size of the river varies 5 m long and 2 m deep. Both springs have been greatly mod- based on the flow regime of the hydroelectric plant, and is

98 Cybium 2009, 33(2) Za n e l l a e t a l . Ecology and diet of Telestes ukliva

Table I. - Sampling of Telestes ukliva in the Cetina River system by station. Water No. of fish weighed, No. of fish captured No. of fish in Station Sampling date temperature measured and for population subsample for diet (°C) released studies study S1 1 Feb. 2000 10.7 1 - - 24 Sep. 1998 13.2 5 S2 19 May 1999 11.3 1 - - 29 Jun. 1999 14.1 9 03 Dec. 1998 12.5 112 21 10 28 Sep. 1999 13.2 7 S3 02 Feb. 2000 10.5 22 04 Apr. 2001 11.0 78 24 Sep. 1998 12.9 39 03 Dec. 1998 12.0 153 15 S4 17 Mar. 1999 11.0 76 10 10 19 May 1999 12.3 6 03 Nov. 2000 11.8 53 10 18 May 1999 12.5 43 29 Jun. 1999 17.6 8 8 19 Apr. 2000 12.4 11 S5 - 12 Jul. 2000 16.9 10 19 Sep. 2000 14.8 3 03 Apr. 2001 10.8 6 24 Sep. 1998 14.2 84 9 03 Dec. 1998 10.3 4 16 Mar. 1999 10.5 2 18 May 1999 13.9 128 10 10 29 Jun. 1999 18.6 13 S6 28 Sep. 1999 15.7 10 02 Feb. 2000 4.8 12 19 Apr. 2000 13.4 10 12 Jul. 2000 18.8 50 19 Sep. 2000 15.5 150 10 10 03 Apr. 2001 12.8 62 S7 18 May 1999 13.3 8 Total 1176 93 40 approx. 25-30 m wide and usually 2-3 m deep. The substrate and S7, the river was relatively deep and so river sampling is pebbles and large rocks. Station S6 (Blato na Cetini) is was conducted by electro-fishing from a boat.T he monitor- further downstream, and takes on more of a lowland appear- ing study included eleven sampling occasions in the period ance with more lentic habitats, and a substrate of small and 1998-2001. A total of 1176 individuals of T. ukliva were col- medium gravel and sand. The river is 20-30 m wide and lected, weighed and measured (total length: TL), and about 0.2-1.3 m deep. The final investigated station was sta- released (Tab. I). To examine the population structure of tion S7 (Radmanove mlinice), where the river depth and ukliva dace, a total of 93 specimens were collected at four width is similar to that at station S5, with a substrate of cob- stations (S3, S4, S5, S6) and were overanaesthetized with bles and pebbles. methane tricainesulfonate (MS-222) and frozen, to prevent any post-mortem digestion of stomach contents for the diet Sampling analysis. In the laboratory, in a subsample of 40 fish (10 per As the ukliva dace is an endemic species and given its season, as follows: from stations S3 and S4 in December current threat status, sampling methods were carefully 1998 and March 1999, from station S6 in May 1999 and selected so as to allow for the return of the majority of fish September 2000), the digestive tract was removed and stom- alive to the river. Sampling was conducted by electro-fishing ach contents stored in 80% ethanol solution. The remaining equipment (using a 2.5 kW stationary DC generator) with specimens (53) were archived in the collection of the Natu- one operator carrying the anode and two assistants holding ral History Museum in Rijeka, Croatia and the collection of hand nets to create a downstream barrier. At stations S2, S5 the Department of Biology, Faculty of Science, University

Cybium 2009, 33(2) 99 Ecology and diet of Telestes ukliva Za n e l l a e t a l .

of Zagreb, Croatia. For population structure studies, all fish 0.08 5.58 0.02 0.22 0.36 0.21 0.43 S7 were preserved using 4% formaldehyde prior to being 93.10 weighed and measured. All individuals were measured for total length (to the nearest mm) and weighed to the nearest 6.43 3.42 S6 32.38 46.85 10.92 0.1 gram. Sex was determined by means of gonad examina- tion. 2.71 0.06 0.89 2.35 S5 93.99 Diet analysis 2.53

0.16 Only gut contents were considered, and were examined S4 97.31 in successive samples under a binocular magnifying lens. foods were counted and identified by reference to 0.04 S3

99.96 parts that commonly survived digestion. The method of fre- quency of occurrence was used for diet analysis. The relative 1.24 0.01 3.05 0.43 4.27 0.41 2.52 S2 importance of animal foods was evaluated by calculating the 88.06 % relative biomass relative % numerical percentage of each prey category (Hyslop, 1980). Prey preferences were displayed by the Ivlev Electivity 0.02 S1 15.30 84.67 Index (Jacobs, 1974) and determined using the following formula: E = (r – p) / (r + p), where r is the relative abun- % 0.01 1.12 0.01 0.03 0.34 0.25 0.06 0.59 0.29 1.17 6.41 total 15.30 56.40 18.04 dance of food category in the digestive tract (as a proportion or percentage of all digestive tract contents) and p is the rela- 6.4 8.2 26.2 44.7 tive abundance of this prey in the environment. Values range 892.8 272.2 198.0 468.0 231.1 935.0 5122.1 total 12231.0 45079.9 14416.0 from -1 to +1, with negative values indicating rejection or mass (g) mass inaccessibility of the prey, zero indicates random feeding, and positive values represent active selection. A subsequent 4.42 3.31 3.33 1.66 4.42 6.63 0.55 S7 75.69 investigation by Lazzaro (1987) showed that true positive or negative prey selection can be interpreted only at values 4.47 0.36 S6 10.88 38.90 45.39 > 0.3 or < -0.3, respectively. To support the feeding studies, food availability was 1.40 1.20 0.40 S5 assessed in May 1999 and September 2000, by sampling the 16.23 80.76 benthic invertebrate communities at station S6, at the same time as the fish were sampled.T hree random samples of bot- 5.49 0.29 S4 94.22 tom fauna were taken by pushing the triangular net (mouth: 20 x 25 cm; mesh 0.25 mm) through the upper centimetres 0.45 S3

99.55 of substratum, and preserved in 4% formaldehyde. In the laboratory, organisms were removed from the substratum, 0.54 0.54 1.08 6.49 1.08 7.57 S2 counted, and identified to the lowest possible taxonomic cat- 72.43 10.27 % relative abundance relative % egory. The number of were converted to 1 m2 of the bottom, and expressed as the numerical percentage of each 3.70 S1 14.81 81.48 prey category. % 0.26 0.20 0.20 0.10 0.03 5.49 4.77 0.03 0.72 2.55 0.13 6.34 38.42 40.77 total RESULTS AND DISCUSSION 8 6 6 3 1 1 4 22 78 168 146 194 n 1176 1248 total Relative abundance, biomass, spatial and temporal distribution Ukliva dace is found throughout most of the river, but it is most abundant in the middle stretch of the river and in the springs on the left tributaries of the Cetina River (Tab. II). In the overall catch during the monitoring program, Telestes ukliva was found to be the second most abundant fish spe- cies in the Cetina River (38.42%), second only to Squalius illyricus (40.77%). However, due to the small size of the Knipowitschia caucasica Knipowitschia Liza aurata Liza canestrini Pomatoschistus fluviatilis Salaria Thymallus thymallus Thymallus anguilla Anguilla Aulopyge huegelii Aulopyge Cyprinus carpio Cyprinus parva Pseudorasbora Cobitis dalmatina Cobitis Carassius gibelio Carassius Telestes ukliva Telestes Squalius illyricus Squalius Salmo trutta Salmo Species

able II. - Percentage of relative abundance and biomass of fish species for seven stations of the Cetina River system. River Cetina the of stations seven for species fish of biomass and abundance relative of Percentage - II. able

T ukliva dace, it accounted for only 6.41% of fish biomass (as

100 Cybium 2009, 33(2) Za n e l l a e t a l . Ecology and diet of Telestes ukliva compared to 56.40% for S. illyricus) (Tab. II). The overall At the springs, water temperature was relatively constant relative abundance of ukliva dace by station was also nega- year round and varied from 11 to 13.2°C, while in the main tively correlated to the overall number of fish species present. river course, temperature ranged from 4.6°C in February to At station S3, ukliva dace was the only fish species recorded, 19.7°C in July. In May, when the ukliva dace is already save for a single specimen of Pseudorasbora parva (Tab. II), abundant in the middle reaches of the river, the water tem- while at station S4, it accounted for 94.22% of all fish, fol- perature is above 12°C (Tab. I). During the late autumn, lowed by Pseudorasbora parva and Salmo trutta. when the water temperature drops to less than 11°C, the An analysis of the catch results from the stations shows ukliva dace retreats back into springs and subterranean that the greatest number of individuals of ukliva dace was waters where the water temperature is stable and higher than captured at station S6. At this site, the river is relatively shal- in the main river course. According to Nikolsky (1963) and low and has several lentic habitats along the left bank. Wocher Schlosser (1991), during winter fishes migrate to deeper and Rösch (2006) emphasized the importance of these habi- waters of the river where there is a lower flow velocity and tats for the related species Telestes souffia(Risso, 1827), as smaller temperature fluctuations. Furthermore, Wocher and the majority of specimens were captured in pools of the Argen Rösch (2006) founded that the related species T. souffia had River in Germany. At stations S2, S5 and S7, only a small a similar migratory behaviour to small tributaries of the river number of ukliva dace were captured, which could be a reflec- that are its wintering habitat. It is likely that the ukliva dace tion of the sampling methods, i.e. due to the width and depth enters the main river course to feed and spawn. Spawning of the river in these reaches, sampling was only possible by occurs in April (unpubl. data) when the temperature exceeds boat, and the appropriate lentic habitats were inaccessible. At 11°C, though the exact spawning locations in the Cetina station S7, the small catch of the ukliva dace could be the River are still unknown. Bless (1996) and Kainz and Goll- result of a relatively dense population of the predatory spe- mann (1990) had similar findings for the related species cies, Anguilla anguilla. In the warmer months, the ukliva dace T. souffia that also spawns at water temperatures between is not found in the springs at stations S3 and S4, with only a 10°C and 12°C. Krivokapić (2002a) also presumed that the few sporadic individuals captured in May 1999 (Tab. I). water temperature threshold for the start of spawning of the The majority of the ukliva dace population exhibited species Telestes montenegrinus (Vuković, 1963) in the migratory behaviour during the year and displayed pro- Morača River is approximately 12°C. nounced temporal distribution. In the warmer part of the The results of the present study allow for the assumption year, ukliva dace inhabits the main river course, while in that water temperature is the most important factor for this winter, specimens colonised springs and subterranean pronounced temporal distribution of the ukliva dace. Tem- waters. In late autumn, specimens migrate and concentrate perature is, likely, a trigger for mass emigration into the primarily in lateral (side) springs of the middle stretch of the springs and subterranean waters. This is in accordance with river, where it is certainly the dominant fish species. Similar Mills (1991), who concluded that water temperature is the behaviour has been observed in different fish species in primary exogenous factor influencing the migratory behav- European rivers, showing seasonal fluctuations in number iour of many cyprinid species. and migrating to different wintering habitats (Nikolsky, 1963; Schlosser, 1991). In December 1998, 265 individuals Population structure and sex ratio were captured in the springs at stations S3 and S4. It is evi- A total of 93 fish were included in the population study. dent that during this part of the year, the ukliva dace is no Males (n = 58) captured ranged in length from 53 to 131 mm longer abundant in the middle course of the river. From mid (TL) and in mass from 1.3 to 25 g, and were smaller than December to early March, the Cetina River is rich in water, females (n = 35), which ranged in length from 71 to 134 mm and the river overflows its banks in many areas. Favourable and in mass from 3.2 to 28.1 g (Tab. III). The Student’s t-test hydrological conditions were only in place during February indicated that the difference in body length between the 2000, when sampling was conducted at all of the investigat- sexes was statistically significant (p < 0.001). For the species ed stations. However, the ukliva dace was captured with only Telestes montenegrinus, females were also found to be larger a single specimen at station S1, in more substantial numbers and have greater mass than males (Krivokapić, 2002b). A at the spring S3 and in small numbers at station S6. These difference in the body size is the most frequent form of sexu- results suggest that during winter, when water temperatures al dimorphism. A large female body size is an advantage for reach their minimums, the majority of the ukliva dace popu- producing more offspring and maximizing reproductive suc- lation retreats into subterranean waters. In mid March or in cess (Nikolsky, 1963; Andersson, 1994). Males and females early April, it is again abundant in the lateral springs (with also exhibited differences in overall length and mass by sta- 76 individuals captured at S4 in March 1999, and 78 indi- tion: the smallest specimens were captured at the Krenica viduals at S3 in April 2001). Later, they again migrate into spring (S3) while the largest were found at the downstream the main course of the Cetina River. station Čikotina lađa (S5) (Tab. III).

Cybium 2009, 33(2) 101 Ecology and diet of Telestes ukliva Za n e l l a e t a l .

Table III. - Mean total length (TL in mm), mass (g) and sex ratio of ukliva dace in population study (n = 93).

Total S3 S4 S5 S6 Males (n) 58 13 21 4 20 TL 85.55 ± 14.50 80.44 ± 9.23 81.05 ±12.99 110.67 ± 10.90 86.5 ± 11.7 g 7.01 ± 3.79 5.17 ± 1.94 5.74 ± 2.75 14.99 ± 5.10 7.11 ± 2.9 Females (n) 35 8 14 4 9 TL 92.84 ± 15.75 88.02 ± 11.36 88.85 ± 12.97 115.38 ± 13.94 91.64 ± 14.03 g 9.21 ± 5.49 7.66 ± 3.80 7.87 ± 4.25 16.21 ± 5.41 9.13 ± 6.04 Sex ratio (M:F) 1.6:1 1.6:1 1.5:1 1:1 2.2:1

Table IV. - Diet composition (% occurrence and % of number of prey categories) of ukliva dace in the Cetina River system. (a) : adult. December March May September % occ % num % occ % num % occ % num % occ % num Aquatic invertebrates Nematoda - 16.7 33.3 - - Gastropoda Ancylus 25.0 20.0 - - - Hydracarina - - - 12.5 1.9 Diptera Simuliidae - - 87.5 83.2 62.5 84.2 Chironomidae 50.0 60.0 16.7 33.3 62.5 10.2 50.0 10.5 Diptera (a) Nematocera 12.5 4.0 - 37.5 2.5 25.0 1.9 Trichoptera 12.5 4.0 - 12.5 0.3 - Ephemeroptera 25.0 12.0 16.7 33.3 25.0 1.1 25.0 1.0 Terrestrial invertebrates Hemiptera Aphidae - - 12.5 1.7 - Cicadae - - 25.0 0.8 - Hymenoptera Chalcoidea - - - 12.5 0.5 Coleoptera Curculionidae - - 12.5 0.3 - Unidentifiable digested matter 50.0 - 66.7 - 12.5 - 50.0 - Total number of food items - 25 - 3 - 363 - 209

The overall sex ratio was 1.6:1 in favour of males. It is number of males than females were captured (Tab. III). known that the sex ratio for most species is near equal, but this may vary within a population, between population and Feeding patterns from species to species (Wooton, 1992). Such variation in Stomach contents were examined in a total of 40 fish, at growth is likely an adaptive response to different ecological the stations where the ukliva dace is most abundant and com- conditions, as females require better environmental condi- mon. Ten stomachs were examined for each season; from the tions than males. One of the primary factors that could influ- springs (stations S3 and S4) in December and March and ence and determine sex ratio is food availability (Nikolsky, from the river course (S6) in May and September, in line 1963; Wootton, 1992). The sex ratio also varied markedly by with the temporal distribution of the fish in the river (Tab. I). sampling location. The sex ratio was 1.6:1 at the Krenica A total of ten stomachs were empty and therefore were not spring (S3) and 1.5:1 at the Jezerce spring (S4), both in taken into account in later analyses. The frequency of empty favour of males. At the midstream station S4, the ratio was stomach by season ranged from 20% (in December, May and 1:1 while at the downstream station S5, more than twice the September) to 40% (March). In the remaining 30 stomachs,

102 Cybium 2009, 33(2) š

Za n e l l a e t a l . Ecology and diet of Telestes ukliva

- 600 prey items were found (Tab. IV). Due to the small Table V. - Relative densities (ind.m ²) of benthic macroinvertebrates at station S6 - Blato na Cetini. number of specimens analysed, only the clearest, general patterns of variation in diet are described. The diet of ukliva May September dace showed marked seasonal variation, and feeding intensi- Tricladida - 2.80 ty is not uniform year round. Benthic invertebrates account- Gastropoda 8.31 17.83 ed for the majority of food share in the analysed stomachs. The ukliva dace consumed few terrestrial insects, ingested Bivalvia - 1.75 only in May and September, periods when ukliva dace are Oligochaeta 1.04 0.35 abundant in the main river course. Despite the preferred len- Hirudinea 0.45 - tic habitat in the middle of the water column and near the Hydracarina 1.34 1.05 surface of the water, the ukliva dace obviously prefers aquat- Amphipoda - 1.05 ic to terrestrial invertebrates. The community composition of Ephemeroptera 32.49 7.69 benthic macroinvertebrates at station S6 is shown in table V. In May and September periods, the presence of certain inver- Plecoptera 1.19 7.34 tebrate groups varied due to their biological and ecological Odonata - 0.35 characteristics. In May, Ephemeroptera, Coleoptera and Coleoptera 22.26 51.05 Simuliidae dominated the macrozoobenthos. In September, Trichoptera 8.46 1.75 the dominant macroinvertebrate groups were Coleoptera and Diptera - - Gastropoda (Tab. V). During the warmer part of the year, food consumption Limonidae 0.45 - was high. Approximately 95% of total prey was found in the Simuliidae 15.58 1.75 guts of fishes captured during May and September periods. Chironomidae 8.46 5.24 During these periods, immature black-flies were the most Total (ind. m-2) 4493.00 1906.00 important food type and highly favoured in both seasons. The Ivlev Electivity index also showed positive selection for Ephemeroptera in both seasons, while Trichoptera and Chi- that ukliva dace fed on benthic macroinvertebrates from the ronomidae larvae were selected only during certain sampling substrate surface and occasionally collected adult insects periods (Tab. VI). Simuliid larvae appeared in the diet of from the water surface. In five guts, some algae (Cyanophyc- ukliva dace more frequently than in the available benthic eae, Chlorophyceae, Xanthophyceae and Diatomeae) were macroinvertebrate populations (Tabs V, VI). The second detected among undigested matter, though these were likely most important prey were Chironomidae larvae, and all head unintentionally ingested along with the macroscopic inverte- capsules found in the guts were identified asCricotopus and brates. Orthocladius. It is well known that the aquatic stages of During December and March periods, specimens of ukli- Simuliidae and Chironomidae are an important source of va dace were caught in springs and black flies were not food for fishes (McLarney et al., 1974, Crosskey, 1990; Rier- present in gut contents. In these periods, feeding activity adevall et al., 1995; Mičetić et al., 2008). Comparisons decreased. In the ten guts collected from the Krenica spring between prey present in the substratum and captured prey (S3), only 25 preys were found, with Chironomidae larvae highlighted the invertebrates avoided by the fish, although and the limpet Ancilus fluviatilis as most abundant. Chirono- high densities were present within the substratum. These midae larvae had a high frequency of occurrence during all were the Coleoptera (mainly Esolus adults) and Gastropoda. seasons, with the exception of March, immediately prior to Similar results were obtained by Cavalli et al. (2003), during the spawning period, when stomachs contained very few an investigation of diet and growth of the endangered Zingel food items (Tab. IV). Due to the increased gonad volume, asper. The rejection of coleopterans in benthos may be due the volume of the digestive tract is greatly reduced and less to their lower energetic value owing to their intense scleroti- food is found in the gastrointestinal tract (Daoulas and sation (Power, 1992). On the other hand, Ephemeroptera Economidis, 1984). Unidentifiable digested matter was nymphs were rarely consumed, despite being very abundant found with high frequency throughout the investigated peri- in the benthic macroinvertebrate community. It is obvious od, with the exception of late spring, when food digestion

Table VI. - Ivlev Electivity Index (E) for benthic macroinvertebrates ingested by ukliva dace at station S6.

Simuliidae Chironomidae Ephemeroptera Trichoptera Hydracarina May 0.68 0.09 0.93 0.93 -1 September 0.96 0.33 0.77 -1 0.01

Cybium 2009, 33(2) 103 Ecology and diet of Telestes ukliva Za n e l l a e t a l . was the highest and unidentifiable digested matter was found CRKVENČIĆ I., DERADO K., FRIGANOVIĆ M., KALOĐERA in only one gut. A., MIRKOVIĆ D., RADICA T., RIĐANOVIĆ J., ROGIĆ V., ROGLIĆ J., STRAŽIČIĆ N. & ŠEGOTA T., 1974. - Geogra- According to these results, the ukliva dace is predomi- phy of Croatia. South Croatian coast. Book 6. 229 p. Zagreb: nantly insectivorous, with aquatic insects as the dominant Školska knjiga. food item. Plant matter was found in 5 stomachs, compris- CROSSKEY R.W., 1990. - The Natural History of Blackflies. ing the clean water algae Cholorphyceae, Cyanophyceae, 711 p. British Museum (Natural History), London. Chichester: John Wiley and Sons. Xanthophyceae and Diatomeae, and fragments of higher DAOULAS C.C. & ECONOMIDIS P.S., 1984. - The feeding of plants. Studies on the diet of Telestes species are rare. Our Rutilus rubilio (Bonaparte) (Pisces, Cyprinidae) in Lake Tri- results are not in agreement with the findings on the diet chonis, Greece. Cybium, 8: 29-38. composition of the endemic T. montenegrinus from the GASTON K.J. & DAVIS R., 1994. - Hotspots across Europe. Bio- Morača River in Montenegro (Krivokapić, 1992), though div. Lett., 2: 108-116. this may possibly be a reflection of the relatively small sam- HYSLOP E.J., 1980. - Stomach contents analysis. A review of ple of T. ukliva guts analysed. T. montenegrinus feeds per- methods and their application. J. Fish Biol., 17: 411-429. manently, year round, with a low incidence of empty diges- JACOBS J., 1974. - Quantitative measurement of food selection. A modification of the forage ratio and Ivlev’s Electivity Index. tive tracts, and showed a broader range of acceptable food. Oecologica, 14: 413-417. In summer and autumn, aquatic invertebrates (with the KAINZ E. & GOLLMANN H.P., 1990. - Beiträge zur Verbreitung exception of black flies) were the dominant prey (79 and einiger Kleinfischarten in österreichischen Fließgewässern. 75% respectively), followed by terrestrial insects (15 and Österr. Fisch., 43: 187-192. 17%) and a small share of algae (6 and 8%). However, dur- KETMAIER V., COBOLLI M., de MATTHAEIS E. & BIANCO P.G., 1998. - Allozymic variability and biogeographic relation- ing winter and spring, when insect food sources are scarce, ships in two Leuciscus species complexes (Cyprinidae) from algae were the dominant item in gut contents (> 50%). In southern Europe, with the rehabilitation of the genus Telestes winter, the second most common component in gut contents Bonaparte. Ital. J. Zool., 65 (Suppl.): 41-48. was detritus (35%), while in spring, aquatic insects were KOTTELAT M. & FREYHOF J., 2007. - Handbook of European Freshwater Fishes. 646 p. Cornol: Kottelat, and Berlin: Frey- common (38%). These differences are likely due to differ- hof. ences in the life cycle and ecological characteristics of its KRIVOKAPIĆ M., 1992. - Nutrition of endemic subspecies Leu- habitat, as T. ukliva spends the colder months of the year in ciscus souffia montenegrinus (Vuković 1963) from the Moraca the springs and subterranean waters, and remains primarily River (Montenegro, Jugoslavia). Glas. Rep. Zav. Zašt. Prirode carnivorous, with a high incidence of empty digestive tracts - Prirod. Muz. Podgorica, 25: 93-103. when food sources are scarce or limited. KRIVOKAPIĆ M., 2002a. - Spawning of subspecies Leuciscus souffia montenegrinus (Cyprinidae, Pisces) from the Moraca River. Nat. Montenegrina, Podgorica, 1: 153-158. Acknowledgements. - We wish to thank Marina Ercegovac for KRIVOKAPIĆ M., 2002b. - Biological study of species of the delta assistance in the laboratory and an anonymous reviewer for helpful of Skadar Lake - Leuciscus souffia montenegrinus (Osteich- comments to the manuscript. We sincerely thank Ivančica Ternjej thyes, Cyprinidae). Meristic parameters. Nat. Montenegrina, for constructive comments. This work was supported by grants Podgorica, 1: 135-152. from Croatian Electric (HEP). LAZZARO X., 1987. - A review of planktivorous fishes: their evo- lution, feeding behaviours, selectivities, and impacts. Hydro- biol., 146: 97-167. REFERENCES MCLARNEY W., HENDERSON S. & SHERMAN M.M., 1974. - A new method for culturing Chironomus tentans Fabricius lar- ANDERSSON M., 1994. - Sexual Selection. 624 p. Princeton, NJ: vae using burlap substrate in fertilized pools. Aquaculture, 4: Princeton Univ. Press. 267-276. BĂNĂRESCU P.M., 2004. - Distribution pattern of the aquatic MIČETIĆ V., BUČAR M., IVKOVIĆ M., PIRIA M., KRULIK I., fauna of the Balkan Peninsula. In: Balkan Biodiversity Pattern MIHOCI I., DELIĆ A. & KUČINIĆ M., 2008. - Feeding ecol- and Process in the European Hotspot (Griffith H.I., Kryštufek ogy of Sabanejewia balcanica and Cobitis elongata in Croatia. B. & Reed J.M., eds), pp. 203-219. Dordrecht: Kluwer Aca- Folia Zool., 57: 1-13. demic Publishers. MILLS C.A., 1991. - Reproduction and life history. In: Cyprinid BLESS R., 1996. - Zum Laichverhalten und zur Ökologie früher Fishes. Systematics, Biology and Exploitation (Winfield I.J. & Jugendstadien des Strömers (Leuciscus souffia Risso, 1826). Nelson J.S., eds), pp. 483-508. London, New York: Chapman Fischökologie, 10: 1-10. & Hall. BONACCI O. & ROJE-BONACCI T., 2003. - The influence of MRAKOVČIĆ M., KEROVEC M., PLENKOVIĆ-MORAJ A., hydroelectrical development on the flow regime of the karstic MIHALJEVIĆ Z., MUSTAFIĆ P., BUKVIĆ-TERNJEJ I., Cetina River. Hydrol. Process., 17: 1-15. RAZLOG-GRLICA J., RADOVIĆ D., KOVAČIĆ D., ĆALETA M., RADIĆ I., ZANELLA D., SCHNEIDER D. & CAVALLI L., PECH N. & CHAPPAZ R., 2003. - Diet and growth GOTTSTEIN-MATOČEC S., 2001. - Evaluation of biological of the endangered Zingel asper in the Durance River. J. Fish resources in the drainage basin of the Cetina River. 241 p. Biol., 63: 460-471. Zagreb: PMF. [In Croatian]

104 Cybium 2009, 33(2) Za n e l l a e t a l . Ecology and diet of Telestes ukliva

MRAKOVČIĆ M., BRIGIĆ A., BUJ I., ĆALETA M., MUSTAFIĆ SMITH K.G. & DARWALL W.R.T., 2006. - The Status and Distri- P. & ZANELLA D., 2006. - Red Book of Freshwater Fish of bution of Freshwater Fish Endemic to the Mediterranean Basin. Croatia. 253 p. Zagreb: Ministry of Culture, State Institute for V+34 p. Gland, Switzerland and Cambridge, UK: IUCN. Nature Protection. WOCHER H. & RÖSCH R., 2006. - Migrations of soufie (Leucis- NIKOLSKY G.V., 1963. - The Ecology of Fishes. 353 p. London: cus souffia agassizii, Val. 1844) in a natural river and a tributary Academic Press. determined by mark-recapture. Arch. Hydrobiol., 165: 77-87. PARKER G.A., 1992. - The evolution of sexual size dimorphism in WOOTTON R.J., 1992. - Fish Ecology. 212 p. Glasgow and fish.J. Fish Biol., 41: 1-20. London: Blackie and Son. POWER G., 1992. - Seasonal growth and diet of juvenile chinook ZANELLA D., MRAKOVČIĆ M., MUSTAFIĆ P. & ĆALETA M., salmon (Oncorhynchus tshawytscha) in demonstration chan- 2008. - Recovery of Telestes ukliva (Heckel, 1843), an endemic nels and the main channel of the Waitaki River, New Zealand. species from the Cetina River, Croatia (, Cyprini- Ecol. Freshw. Fish, 1: 12-25. dae). J. Fish Biol., 73(1): 311-316. RIERADEVALL M., GARCIA-BERTHOU E. & PRAT N., 1995. - Chironomids in the diet of fish in Lake Banyoles (Catalonia, Spain). Chironomids. In: From Genes to Ecosystems (Cranston P., ed.) pp. 335-340. East Melbourne: CSIRO Publications. SCHLOSSER I.J., 1991. - Stream fish ecology: A landscape per- spective. BioScience 41: 704-712. SKET B., PARAGAMIAN K. & TRONTELJ P., 2004. - A census of the obligate subterranean fauna of the Balkan Peninsula. In: Balkan Biodiversity Pattern and Process in the European Hotspot (Griffith H.I., Kryštufek B. & Reed J.M., eds). pp. 309- Reçu le 25 septembre 2008. 322. Dordrecht: Kluwer Academic Publishers. Accepté pour publication le 13 mai 2009.

Cybium 2009, 33(2) 105