Food of bream, Abramis brama [L.] and white bream, Blicca bjoerkna [L.] in Zegrzynski Dam Reservoir

B ream and white bream inhabit inland water bodies lowland rivers and marine lagoons. The two species are very similar morphometrically and are characterized by similar feeding behaviour (Brabrandt 1984, Lammens 1982, Lammens 1984, Rask 1989, Wielgosz, Tadajewska 1988). Along with increasing eutrophic:ation the two species become most important components of the fish stock biomass in the water bodies. Polish literature does not have many papers on these two fish species in dam reservoirs (Klimczyk-Janikowska 1974, Martyniak et al., 1987, Wielgosz, Tada· jewska 1988), lakes (Prejs 1973, Budzyn.ska et al., 1956) or rivers (Pliszka 1981, Terlecki et al., 1977). This paper represents a part of complex studies on feeding relations in fish of Zegrzyii.ski Dam Reservoir (the project was financed by the Institute of Ecology of the PA S). The aim of the study was determine the diet of bream and white bream at three stations in Zegrzyii.ski Dam Reservoir. Magdalena TADAJEWSKA

Feeding conditions for the zooplankton are good due to high phytoplankton bio· mass, constant inflow of organic matter to the reservoir and low pressure of inverte· brate predators (Simm 1990, Ejsmont-Karabin, Wf;glenska 1990). Average biomass of rotifers and crustaceans is the highest in the central part of the reservoir, with the domination of crustaceans, on the average 1.86 mg dm·3, followed by the part of former Bug River (with the domination of crustaceans, 0.48 mg dm" 3 on the average).
The lowest biomass was noted in the former Narew River (with the domination of crustaceans, 0.23 mg dm" 3 on the average). Maximal concentrations of the zoo· plankton in this reservoir were comparable to zooplankton densities in hypertrophic lakes (Ejsmont-Karabin, W�glenska 1990). Also numbers and biomass of benthic organisms reached high values. The main taxonomic groups were Mollusca, Oli gochaeta and Chironomidae, and within the latter Chironomus f.l. plumosus, Pro cladius sp. and Glyptotendipes e.g. gripekoveni (Dusoge et al. 1990, Kuklifiska 1989. Numbers of Chironomidae larvae were especially high in the first part of the vegetation season (May-June). In the Bug River part of the reservoir Chironomidae biomass reached 850 g m-2 , over 70 thousand individuals (Kuklifiska 1989). Mollusca were more numerous in the part with water flow compared to the still water part (Dusoge et al. 1990). It should be mentioned that thermal conditions and water inflow differed in parti cular years. In 1988 the vegetation season was longer; there were also more days with 20°C than in 1987. Water inflow in 1988 was higher than in 1987 (9.955 x 10'6 cubic m, 7.375 x 10'6) (Kajak 1990 b). Bream and white bream as well roach were dominating in the experimental catches (Grudniewski, Boron 1990) and in the commercial ones.
In view of different environmental conditions in particular parts of the reservoir, sampling stations were selected for the two fish species under study. Fish were caught on three fishing grounds. The same stations were used in the studies on bottom fauna (KukHfiska 1989). a) Wierzbica -located on the right bank of the former Narew River, up the bridge on the road Serock-Wyszkc r w. Bottom sediments up to 1 m were sandy in this place, while deeper layers contained detritus and a few cm loose mud (Kuklinska 1989). This station was characterized by the lowest biomass of phyto-and zooplankton as benthos (Kajak 1990a, Simm 1990, Ejsmont-Karabin, W�glenska 1990, Dusoge et al. 1990. b) Bug -located at the left bank of the former Bug River. This station was charak· terized by steep sloping of the bottom and no submerged vegetation. Bottom at the depth of 1-2 m was covered with loamy layer of mud of several cm (Kuklinska 1989). This station was very fertile, with average biomass of photo-and zooplankton and the highest density and biomass of "soft" bottom fauna (Kajak 1990a, Simm 1990, Ejsmont-Karabin, W�glenska 1990, Dusoge et al. 1990, Kuklinska 1989. c) Zegrze -located in the south part of the reservoir, at the height of Zegrze village. Bottom at the depth of 1-2 m is sandy, covered with a few cm of mud, with remnants of shells, mostly of Dreissena polymorpha and Viviparus sp. This station was characterized by average fertility, the lowest water flow and average of zoo plankton and benthos biomass (Ejsmont-Karabin, W�glefiska 1990, Dusoge et al. 1990, Kuklifiska 1989.

MATERIAL AND METHODS
Materials were collected in 1987 and 1988. Samples were taken every month since April till October. Fish were caught with gill nets of size 40 to 160 mm. Fish were weighed up to 10 g and body lenght was measured (longitudo corporis), 636 food tracs of bream were collected (Tab. 1) and 537 of white bream (Tab. 2).
It was not possible to collect materials of the two species at an stations each month. Collected food tracs were preserved in 4% formalin. Content of each food tract was viewed under a stereoscope microscope determining bigger food components. Chironomidae which dominated in the food were determined to the lowest possible systematic category. Samples were taken from each food tract for determination of Oligochaeta bistles. Mass of eaten components was calculated from the formula W = K L 3 (Morduchaj-Boltovskoj 1954). For most Chironomidae larvae K = 3.5; only for Procladius sp. and Cryptochironomus sp. K = 7. Molluscs in the food tracs were damaged. They were very smashed by pharyngeal teeth. In most cases it was impossible to determine species of these food components so they were determined as Mollusca. Their biomass (together with crustaceans) was calculated using a percentage method (volume-point) from known food weight.
Apart from the main food components items from other systematic groups were also found in the food tracs, as also non-calculable components (plants); they are presented in Table 3. Rare prey organisms (and sometimes their parts) belonged to such systematic units as Cladocera, Ephemeroptera, Isopoda,. Arachnida, Cestoda and Oligochaeta. Since they were sporadic their biomass was not calculated, and their occurance was denoted as "+" (Tab. 3).
In order to facilitate interpretation of the results, samples from each month was treated as a whole for each species (it was not divided into size classes).
In addi t ion to the biomass calculations were also made of the numbers, frequency of occurrence and indices of filling. Biomass of the food consumed was divided into the following prey categories: Chironomidae larvae, Chironomidae pupae, Mollu sca and Ceratopogonidae, and other larvae (eg. Sia/is sp.). In the case of the share in numbers all food components were divided into the following categories: Chironomus sp. larvae, Glyptotendipes sp. and other Chironomidae, non-identi· fied Chironomidae and Chironomidae pupae. Determinations of the importance of particular food components in the diet they were divided into the following groups: eudominants, when the component represented 50.1-100%, dominants from 20.l to 50% subdominants from 5.1 to 20%, rare organisms 1.1-5%, and sporadic compo nents below 1 % of tbe biomass or numbers.

L Bream
Characteristics of bream under study are given in Table 1. Apart from the dates of sampling and range of body lengths the table gives per cent of full food tracs. Index of filling with food is presented in Fig. 2. The highest per cent of full food tracs was found in Zegrze station in May 1987 (91%) and in September of the same year at the same station (83%). The lowest values of this index were found in September 1988 at station Wierzbica (6%) and in October 1988 at station Zegrze (7%). The highest index of filling was found in July 1987 at station Wierzbica and Bug (49% .. ) the lowest (7% .. ) was observed in April 1987 at station Wierzbica (Fig. 2).
Chironomidae larvae were the main components of bream diet. Chironomus sp. larvae represented main components of food biomass both in 1987 (76.35%) and 1988 (64.45%). Glyptotendipes larvae were on the second place in bream diet. In 1988 they were more numerous than in 1987 and they represented 30.6% of the biomass (Tab. 3). On station Wierzbica molluscs were also consumed.

St a t i o n W i e r z.b i c a
No data were collected on this station for two months in 1987 (June and Septem ber). Bream diet in April, July, August and October consisted of Chironomidae larvae. Only in May Mollusca were the main component, while Chironomidae larvae be came a subdominant component (12.8% of the biomass) (Fig. 3). Chironomus sp. larvae were the most frequent component (52%); Glyptotendipes sp. larvae were present as dominants (36.6%) (Fig. 4). Chironomus sp. larvae were present as an eudominant in the samples collected in April (60.8%), July (80.7%) and October (93. 7%) (Fig. 4), Glyptotendipes sp. larvae were present as dominants in the samples collected in August (47.5%) and May (36.6%). Larvae of other Chironomidae were present as dominants in April (21.4%). Larvae of unidentified Chironomidae were present as subdominants in the samples collected in April (17.8%), July (16.9%) and August (15.8%) (Fig. 4).
No data for bream were obtained in April and August 1988 at Wierzbica station. In May and September Mollusca were an eudominant in bream diet, In June, July and October bream diet consisted mainly of Chironomus sp. larvae {47%) and larvae of other Chironomidae which constituted the dominants (45%) (Fig. 4).
,,..  At this station no data were collected for bream in April, May and August 1987, Chironomidae larvae were the main component of the diet They were eudominants in June, September and October {Fig. 4).
In 1988 bream samples were collected on this station in all months (April-October). Chironomidae larvae were the main food components in all cases (Fig. 3). Chironomus sp. and Glyptotendipes sp. larvae were most frequent (Fig. 4). Mollusca were found only in April as an infrequent components (less than 5% of the biomass). Pupae were present in some months: May, June, July and Septem ber. They were most numerous in May, 16.5% of the food biomass (subdominant) (Fig. 4). In the other months {June, July and September) they were infrequent (up to 5% of the biomass) (Fig. 3).
Ceratopogonidae larvae appeared in May sample but they were infrequent, only 4% of the biomass.
Larvae of unidentified Chironomidae were present as subdominants in May (15.7%).

S t a t i o n Ze g rze
Only bream samples were collected from this station. In 1987 no sample was collected in August. Chironomidae larvae were the main component. Pupae appeared in 4 months only: April, May, June and July. Pupae were least frequent in April (0.4% of the food biomass) and most numerous in May (8.8%) (Fig. 3). Chironomus sp. larvae were an eudominant in all months (Fig. 4). In April larvae of other Chirono midae were the dominant (21.3%), in May their place was taken by pupae (26.2%), and in September by Glyptotendipes sp. larvae (30.5%).
Larvae of unidentified Chironomidae were present as subdominants in April (13.4%) and July (14%). In October larvae of other Chironomidae appeared in bream diet as the subdominants (8%).

In June and October Chironomus sp. larvae were the eudominant. Larvae of other
Chironomidae were also found. In June they were present in minimal amounts (1% of the numbers) and in October they became a subdominant (6.3%) in bream diet (Fig. 4).

II. White bream
In the case of white bream most full food tracs were found in April 1987 (79%) and 1988 (75%) at station Bug (Tab. 2). The highest number of empty food tracs was observed in September 1988 at station Wierzbica. The highest index of filling was noted at this station in the same year (85.3% .. ). The lowest value of the index of filling was found at station Bug in October 1987 (5.7% .. ) (Fig. 5).

Stat i o n W i e r z b ica
In 1987 data for white bream were collected only in April, May and June. Chirono midae larvae and Mollusca were main components of white bream diet. In April and May this fish fed only on Mollusca. In June white bream fed on Chironomidae larvae and pupae (Fig. 6). Chironomus sp. larvae decisively dominated in the diet · (eudominant) and unidentified Chironomidae larvae and pupae were present as infrequent components (less than 5% of the number) (Fig. 7).
In 1988 no white bream samples were collected in April, July and August, Mollusca were present in diet in May and October. In June and September white bream food was composed of Chironomidae larvae (Fig. 6). Chironomus sp. larvae represented 100% of the food biomass (Fig. 7).

St a t i o n B u g
Chironomidae larvae and pupae constituted white bream food. In 1987 no data were collected in August, Chironomus sp. larvae were the main food component.

DISCUSSION
Mature bream feeds mostly on invertebrate fauna (Prejs 1973, Wielgosz a. al. Tadajewska 1988. Gastropoda, Bivalvia, Crustacea and Insecta are the main food components (Rask 1989, Stolarov 1985, Lammens a. al., 1987. From among these Chironomidae larvae are decisively the most important. (Ziteneva 1960, Zadoroznaja 1977, Martyniak a. al. 1987. The same was observed in Zegrzynski Dam Reservoir. Chironomidae larvae were the main food items in bream and white bream diet Numbers and biomass of bottom fauna in this resevoir reached very high values. Molluscs predominated -they constituted 93% of the total biomass of bottom fauna in the reservoir. However, it should be noted that shells represent 50% of mollusc biomass, while digestible parts only 46.5%. Apart from the molluscs, bottom fauna contained large numbers of Chironomaidae. They represented 35% of the whole pro· duction biomass (Kajak 1990 a). High numbers were also reached by Oligochaeta, but they were mostly juvenile forms so their biomass was not very high. As regards Chironomidae, Chironomus sp. larvae attained considerable size (up to 25 mm) so their biomass was also high. This was the reason for high Chironomidae biomass (Dusoge a. al. 1990). As results from the studies on zoobenthos (Kuklinska 1989) stations Wierzbica was characterized by the lowest numbers and biomass of Mollusca and of the so-called "soft benthos". Intermediate values were observed on station Zegrze. The highest numbers and biomass of Mollusca, 0/igochaeta and Chirono· midae were noted on station Bug. This station was the most fertile. Biomass of Chironomidae larvae reached 349 gm 2 in 1987 at the depth of l m, Since the larvae were present in high densities, they constituted the main food item in diet of bream and white bream. This station was also very rich in 0/igochaeta, but the two fish did not penetrate deeper mud layers with Oligochaeta (Kuklinska 1989) due to the abundance of Chironomidae larvae. Even roach fed on Chironomidae larvae on this station despite the fact that it is a typical mollusc feeder. These larvae represented 13-72% of the content of roach food tracs (Szczyglinska 1991 (Stolarov 1985). Molluscs were also the main food item of bream in Tvarminne, north part of the Baltic Sea (Rask 1989). Different results were obtained by Lammens (1984) who studies bream diet in eutro phic lakeso When zooplankton was very abundant all size-classes of bream fed on zooplankton, but when its abundance decreased small bream (less than 20 cm) consu med benthic Cladocera while bigger fish -Chironomidae larvae. This autor stated that the amount of food consumed depended on food abundance, size of food orga· nisms and their availability. Zeltenkova and Kogan (1985) stated that benthos can be grazed upon at 55-89% of its production. Hence, possibility of utilizing the food resources depends on the type of prey organisms and their density in a reservoir. Maximal densities of animals are always noted in deeper sediments (Gusar, Pjanov 1988). Eggers (1977) and Ziteneva (1983) proved that total consumption and diet corn· position depended on prey densities.
When there no bottom fauna available bream consumes zooplankton (Zadorozna ja, Sappo 1974). Planktonic crustaceans (Daphnidae) constituted the main compo nent in the diet of mature bream from Lake Tjeukemeer in Denmark (Lammens et al. 1985). This food was sufficient to ensure good fish condition.
Oligochaeta were always estimated as less frequent in feeding studies because they do not possess shells, chytinous caroaces etc. Hence they can be completely digested within an hour after ingestion (Galinksky, Nikitin 1972). In Norwegian lakes Oligochaeta were the main food item in the diet of Coregonidae {Milbrink 1973 b )o Considerable role played by Obligochaeta in the diet of benthic feeding fish was underlined by Mirosnic:enko (1978), Ermolin (1979) and Ziteneva (1982). When Chironomidae numbers decrease, 0/igchaeta become quite important. In Kujby szewski and Cymlaflski reservoirs 0/igochaeta were of primary importance in the diet (Jegereva 1960, Ziteneva 1981. In the recent years detailed studies were performed on the feeding of plankton -eating fish. Two methods of feeding were distinguished: filtration (Eggers 1977, Lammens 1985, Hoogenboezem et al. 1991 and individual feeding. In the case of filtration white bream is less effective than bream. This is due to the structure of gills (Hoogenboezem 1990). When prey organisms are small bream feeds more effecti vely than white bream. But when the food resources are composed of bigger orga nisms and the particles of consumed substrate are bigger, white bream is more effi· cient in filtrating the food than bream (Lammens 1984). Detailed studies were also performed on mouth (Winfield et aL 1983) and the structure of filtrating-gill appa ratus (Hoogenboezem et al. 1990). Lammens et aL (1987) described differences in filtrating processi, pharyngeal teeth and mouth proportions between roach, bream and white bream which allowed these-species to occupy different food niches in the same conditions in eutrophic Danish lakes. Bream is able to penetrate deeper into the bottom sediments because its mouth is more protruding. This fact was observed by Brabrand (1984) when he compared bream with white bream. Roach is the only fish with pharyngeal teeth strong enough to crush mollusc shells (Rask 1989).
Distribution of resources in a water body is a commonly know phenomenon. It is thought that it is a feature of species coexistanceo Distribution of the resources differs depending on utilizing abilities of the coexisting species. Studies showed that overlapping of the diets was very unstable. In the cycles rich in food overlapping was considerable, in poor cycles it was minimal. Lammens et al. (1985) showed that when food resources were abundant and fish were in good condition diets did overlap. When condition decteriorated and food resources decreased there was a change of the food niche and fish diets did not overlap.
It seems that in Zegrzynski Reservoir bream and white bream are coexisting rather than competingo CONCLUSIONS 1. Chironomidae larvae and molluscs were the main food item on station Wierz· bica, Mollusca were present as eudominants in bream diet in May and Septem· ber, and in white bream diet in April, May and October.

Chironomidae larvae were the main food item on station Bug. Only in May
1988 Chironomidae pupae were present as an eudominant in white bream diet 3. Chironomidae larvae were the main food item in bream diet on station Zegrze.