FOOD SELECTION OF PELED LARVAE (COREGONUS PELED GMEL.), REARED IN ILLUMINATED CAGES IN DIFFERENT WATER BODIES

Grażyna Furgała-Selezniow, Andrzej Mamcarz, Andrzej Skrzypczak
Department of Lake and River Fisheries, University of Warmia and Mazury in Olsztyn, Poland

ABSTRACT

The aim of the experiment was to examine prey size selection of peled (Coregonus peled Gmel.) larvae using Strauss linear index for food selection. For this purpose larval peled were held in illuminated cages at two stocking densities (10 000 and 40 000 individuals per cage) in three water bodies of diverse trophic characteristics. In the first days of rearing peled larvae fed on zooplankton up to 0.5 mm in length. Then most peled larvae significantly more often selected organisms up to 1 mm, which were most abundant in the environment. Under good feeding conditions, reared fish selected medium-sized zooplankters (from 0.6-1.0 mm length class, less often those 1.0-1.5 mm in length). Stronger food pressure for small prey appeared at a time of poor feeding conditions. Fish fed rather randomly on larger planktonic organisms (>1.5 mm), which were not abundant in the environment. Statistically significant differences were observed in food selection between the cages with different stocking densities, even when the feeding preferences were similar. Large variability in the feeding behaviour of peled larvae was observed during the whole rearing period.

Key words: peled larvae, food selection, Strauss linear index of food selection, cage rearing..

INTRODUCTION

Larval food preferences are related to rapid developmental changes in their morphological and anatomic features, alimentary tract, vision and swimming abilities [12, 24], which all result in improved feeding success [1]. Factors affecting selective feeding patterns of larval fish and numbers of organisms consumed include prey size [10], prey visibility contrast [20], prey and predator encounter rate, prey escape response and prey swimming pattern [2, 5].

As coregonids are of importance to commercial lake management in Poland, a technology of larval rearing in illuminated cages has been introduced [16]. Final results of cage rearing in water ecosystems are highly variable because of seasonal zooplankton fluctuations [3, 13]. Former experiments indicate that larval feeding strategy is adaptable to specific conditions of each lake. The following events have been noted: stock stratification, individual food specialization, changes of feeding time, feeding on atypical prey and even cannibalism at times of prey shortage in illuminated cages [13]. Food preferences of fish larvae for various groups of zooplankton alter as they grow larger [15, 22].

The aim of this experiment was to analyze food selection of peled larvae (Coregonus peled Gmel.) reared in illuminated cages in water bodies different in the trophic status, and to determine dissimilarities in feeding selectivity between two stocking variants. The analysis involved the size of zooplankton.

MATERIALS AND METHODS

Fish farming conditions

Fish rearing in illuminated cages was carried out in 1986-1988, in three water bodies of different trophic characteristics: a fish pond (29.04 - 30.06.1986), Lake Mutek (29.04 - 30.06.1987) and Lake Legińskie (21.04 - 27.06.1988). These three water reservoirs are located close to each other, in the group of the Legińskie lakes (north-eastern Poland). Lake Legińskie, the biggest of the three water bodies (228.3 ha, max. depth 37.2 m), is mesotrophic to eutrophic. Lake Mutek is small (10.7 ha, max. depth 17.2 m), polytrophic and bradymictic. The fish pond, lying ca 50 m off the shore of Lake Legińskie, is small (0.2 ha, max. depth 2 m). Another difference between the three water reservoirs was the time of water warming in spring, which in turn affected trophic conditions and ecological processes in the pond and the lakes [14].

Field observations embraced the larval and early juveniles period of reared fish. The cages were stocked with 10 000 (variant I) and 40 000 (variant II) 1 to 2-day-old peled larvae (5376 and 21 505 fish m^-3). The fish were restricted to the epilimnion (to 3.5 m below water surface) in 1.86 m³ surface cages (mesh size 1.2 mm), replaced after a fortnight with cages of the capacity of 4.5 m³ [16]. Each cage was illuminated with a bulb (60 W / 24 V) submerged at 1.5 m in order to attract zooplankton as fish food. Platforms with the cages were anchored in different zones of these waters. In Lake Legińskie they were placed in the pelagic zone, in Lake Mutek they were put in the littoral zone, and in the pond they were situated in the main ditch, near the bottom.

Methods

In order to determine accessibility of food to reared larvae, samples of zooplankton were collected from an empty cage of the same volume, located directly beside the cages stocked with fish. It was assumed that zooplankton attraction by light in all cages was of the same character [17]. The samples were collected weekly at midnight by a vertical plankton net (mesh size 50 μm) from 1.5 m depth to the water surface [9]. Zooplankton were preserved in 4% formaldehyde, then examined and scored under a microscope. The body size was measured (+/-0.1 mm) for all or at least 30 specimens from each species (except Rotatoria). Plankton organisms were divided into five length classes (every 0.5 mm), from 0.0-0.5 mm to individuals bigger than 2 mm. Rotatoria were classified under the smallest group.

Samples of n = 50 fish were taken from cages weekly at midnight (at the same time as zooplankton) and preserved in 4% formaldehyde. All sampled fish were measured (TL +/-1.0 mm). The alimentary tract was dissected from each specimen and its content was examined under a microscope for prey composition and number. Food organisms were measured and divided into length classes as in the zooplankton samples. One thousand of fish from the pond, 900 from Lake Mutek and 1000 from Lake Legińskie were examined.

Food selection of larvae according to the zooplankton length classes was assessed using a linear index of food selection [21]: L = r_i - p_i , where r_i is the proportion of a food size category in the gut and p_i is the percentage of the same food size category in the zooplankton sample. The Strauss food selectivity index was chosen because of its earlier use with coregonid larvae and well-defined statistical properties [21, 22, 23]. The value of an index ranges from -100 (a negative selection) to +100 (a positive selection). The expected value of L is zero for random feeding and we tested whether L differed significantly from zero with a t-test at 5% significance level. Should a given group of zooplankton be present in the environment but not ingested by fish, test (t) statistics and standard deviation (SD) values equaled zero. When a certain group was absent from both the zooplankton and the diet, the food selectivity index value could not be calculated. Differences in feeding preferences between both stocking density variants were also tested (t-test, p<0.05).

RESULTS

Analysis of zooplankton

Cage zooplankton density in the pond ranged from 389 indiv. l ^-1 on day 35 to 6189 indiv. l ^-1 on day 42 of experiment (Fig. 1). Organisms shorter than 0.5 mm (rotifers, juvenile copepods, and small cladocerans) prevailed in the plankton, amounting to 38.8-92% of the total number of organisms (Fig. 2). It was only on day 21 of rearing that the 0.6 to 1.0 mm length class was the most abundant (53.6%). At the beginning of the experiment specimens of Cladocera (mainly Daphnia sp.) measuring 1.1 to 1.5 mm were numerous (30.7%). The maximal percentage of the zooplankton groups longer than 1.5 mm was 1.9%.

Fig. 1. Cage zooplankton density in three reservoirs during the rearing

In Lake Mutek the highest concentration of planktonic organisms, 3543 indiv. l ^-1, was noted in May, on day 13 of fish rearing (Fig.1). The lowest concentration occurred in early and late June (below 600 indiv. l ^-1). Like in the pond, many of the plankton organisms (28.9-97.7%) were small specimens to 0.5 mm length (Fig. 2). Organisms from the 0.6-1.0 mm and 1.1-1.5 mm length classes were numerous only in the first days of the experiment (49.4 and 21.7%, respectively). In the following weeks their numbers were much lower (mainly ca. 20%). Specimens larger than 1.5 mm (Leptodora kindtii Focke) were observed sporadically.

Fig. 2. Numerical composition of the zooplankton in the illuminated cages. A - the pond, B - Lake Mutek, C - Lake Legińskie

The density of zooplankton in Lake Legińskie ranged from 133 indiv. l ^-1 (mainly Rotatoria) at the beginning of rearing to 3259 indiv. l ^-1 on day 26 of the experiment (Fig. 1). Organisms from the smallest group were the most common zooplankters (70.3-99.9%; Fig. 2). The 0.6-1.0 mm group was less numerous (to 28.3%). Specimens larger than 1 mm were scarce (to 1.5%).

Selection of food organisms

The average length of larval peled at the beginning of cage rearing was from 9.3 to 10.0 mm (Fig. 3). In all the water bodies, especially during the first days of rearing, fish with empty alimentary tracts were found. In the pond and in Lake Mutek, the highest number of fish larvae without food in the gut was recorded in the first samples (in the pond: 8% for density variant I and 56% for variant II on day 2 of rearing; in Lake Mutek: 50 and 58%, respectively, on day 6). In Lake Legińskie, on the other hand, notable ratios of fish with empty alimentary tracts were found for as long as the first three weeks of rearing (96 and 100%, 40 and 88%, 10 and 4% in the lower and higher density variant, respectively), which was due to the poor feeding conditions (Fig. 1). Later on, fish with empty alimentary tracts were recorded only sporadically.

Fig. 3. Growth in length (TL +/-SD) of Coregonus peled larvae reared in illuminated cages in two stocking variants (I and II; see: Material and methods). A - the pond, B - Lake Mutek, C - Lake Legińskie

In the first week of the experiment peled larvae reared in the pond chose the smallest plankton organisms, measuring up to 0.5 mm in length (Table 1). In the successive weeks larval peled tended to prefer organisms from the 0.6 to 1.0 mm length class. The analysis of Strauss food selectivity index reveals that the importance of the 0.6 to 1.0 mm class plankton varied in time and was inconsistent. Between days 7 and 21 it was the most preferred group of plankton organisms. From day 28 on, as the amounts of zooplankton in the environment were declining (Fig. 1) while the ratio of organisms up to 0.5 mm in length was on the increase (Fig. 2), fish began to prefer again the smallest organisms (Table 1). The values of the selectivity index suggest that zooplankton specimens above 1 mm in length were rarely under feeding pressure. Fish ingested such large organisms at random; more frequently they would avoid them in a manner that was statistically significant.

Table 1. The mean values of Strauss food selection index (L) of Coregonus peled larvae reared in illuminated cages in two stocking variants in the pond and the results of comparisons of L values between stocking variants

Days of rearing	Plankton length classes (mm)	I			II			Comparison of I and II
		t1	SD	L	t1	SD	L	t2
2	0.0-0.5 0.6-1.0 1.1-1.5	7.07* -5.92* 0.00	±32.65 ±14.27 0.00	+32.66* -11.96* -30.70	-1.49 0.00 0.00	±50.14 0.00 0.00	-10.60 -14.70 -30.70	-5.11* -1.36 0.00
7	0.0-0.5 0.6-1.0 1.1-1.5 1.6-2.0	0.69 0.37 -47.81* 0.00	±23.47 ±15.24 ±0.73 0.00	+2.28 +0.79 -4.95* -1.90	-0.95 4.48* 0.00 0.00	±12.73 ±12.73 0.00 0.00	-1.71 +8.71* -5.10 -1.90	-1.06 -2.82* -1.43 0.00
14	0.0-0.5 0.6-1.0 1.1-1.5 1.6-2.0 >2.0	-10.15* 13.86* -1.97 -12.65* 0.00	±16.57 ±14.18 ±9.17 ±0.75 0.00	-23.80* +27.79* -2.55 -1.35* -0.10	-3.95* 9.06* -9.27* 0.00 0.00	±18.31 ±16.32 ±6.93 0.00 0.00	-10.23* +20.91* -9.08* -1.50 -0.10	3.88* -2.25* -4.02* -1.43 0.00
21	0.0-0.5 0.6-1.0 1.1-1.5 >2.0	-12.48* 8.88* 4.34* 0.00	±11.77 ±12.08 ±8.20 0.00	-20.77* +15.17* +5.04* -0.10	-1.82 2.31* -16.44* 0.00	±26.42 ±28.03 ±2.69 0.00	-6.80 +9.15* -6.25* -0.10	3.41* -1.40 -9.24* 0.00
28	0.0-0.5 0.6-1.0 1.1-1.5 1.6-2.0 >2.0	5.75* -1.64 -19.93* 0.00 -7.62*	±20.59 ±18.93 ±3.72 0.00 ±0.41	+16.74* -4.40 -10.49* -1.40 -0.44*	0.20 0.66 -10.48* 0.00 0.00	±32.08 ±24.22 ±6.26 0.00 0.00	+0.90 +2.28 -9.27* -1.40 -0.50	-2.94* 1.54 1.18 0.00 -1.00
35	0.0-0.5 0.6-1.0 1.1-1.5 1.6-2.0 >2.0	3.28* 2.56* -17.87* 0.00 0.00	±19.99 ±16.13 ±5.66 0.00 0.00	+9.28* +5.84* -14.31* -0.20 -0.60	2.00 2.25* -12.41* 1.43 0.00	±24.95 ±18.51 ±7.28 +/-2.17 0.00	+7.04 +5.90* -12.78* +0.44 -0.60	-0.49 0.02 1.17 2.08* 0.00
42	0.0-0.5 0.6-1.0 1.1-1.5	-2.21* -7.97* 0.00	+/-14.99 ±5.49 0.00	+4.69* -6.19* -0.50	19.72* -18.37* 0.00	+/-2.62 ±2.62 0.00	+7.29* -6.79* -0.50	1.21 -0.70 0.00
49	0.0-0.5 0.6-1.0 1.1-1.5 1.6-2.0	-2.36* 2.15* 2.84* 1.27	±17.80 ±16.42 ±2.59 ±0.78	-6.43* +5.59* +0.55* +0.28	-3.72* 3.74* 1.00 -2.56*	±19.58 ±19.57 ±0.11 ±0.18	-10.31* +10.36* +0.02 -0.07*	-1.00 1.26 -2.75* -1.56
56	0.0-0.5 0.6-1.0 1.6-2.0 >2.0	22.27* -24.26* -0.21 0.00	±2.67 ±2.43 ±1.09 0.00	+8.43* -8.35* -0.03 -0.10	16.09* -15.51* -1.63 0.00	±3.69 ±3.69 ±0.54 0.00	+8.39* -8.09* -0.12 -0.10	-0.06 0.42 -1.12 -0.54
63	0.0-0.5 0.6-1.0 1.1-1.5 1.6-2.0 >2.0	-2.71* 1.76 1.22 1.75 0.00	±25.11 ±14.42 ±1.25 ±7.80 0.00	-9.63* +3.58 +0.22 +1.94 -0.10	1.17 -4.27* -10.54* -6.43* 0.00	±15.00 ±6.54 ±0.18 ±0.17 0.00	+2.48 -3.95* -0.27* -0.16* -0.10	2.93* -3.36* -2.75* -1.89 0.00

I and II; see: Material and methods. Indexed (*) t test statistics values are statistically significant (p<0.05). Indexed (*) L values differed significantly from zero (p<0.05). 1 Critical value t = |2.01| (df 49; α=0.05). 2 Critical value t = |1.98| (df 98; α=0.05).

While rearing peled larvae in the pond, some statistically significant differences in the feeding preferences for single zooplankton groups were observed between cages with different stocking densities (Table 1). It was only on days 14, 21 and the last day of the experiment that such differences involved more than one class of zooplankton. On days 14 and 21, although the fish from both stocking densities I and II chose organisms from the class of 0.6 to 1.0 mm and avoided those belonging to the 0.0 to 0.5 mm class, this tendency manifested itself more clearly in the cage with the lower density, as indicated by the values of the selectivity index (L) and test (t) statistics. On the last day of rearing, fish from the lower stocking density cage neglected the smallest organisms, while those from the higher density cage avoided the zooplankton larger than 0.6 mm. The average length of a peled larva was 34.1 mm in cage I and 31.8 mm in cage II (Fig. 3).

In Lake Mutek, same as in the pond, those fish which had begun foraging chose organisms from the 0.0 to 0.5 mm class (Table 2). Starting from day 13, with an exception of day 20 in both cages and day 41 in the higher density cage, larvae ingested larger prey, especially those from the 0.6 to 1.0 mm class (the highest values of Strauss linear food selection index, significantly different from zero), even though the most numerous in the environment were the zooplankton from the 0.0-0.5 class (Fig. 2). Positive selection of the zooplankton above 1 mm in length was also observed, particularly in the last three weeks of the experiment.

Table 2. The mean values of Strauss food selection index (L) of Coregonus peled larvae reared in illuminated cages in two stocking variants in Lake Mutek and the results of comparisons of L values between stocking variants in Lake Mutek and the results of comparisons of L values between stocking variants

Days of rearing	Plankton length classes (mm)	I			II			Comparison of I and II
		t1	SD	L	t1	SD	L	t2
6	0.0-0.5 0.6-1.0 1.1-1.5	-2.21* -118.09* 0.00	±37.26 ±2.91 0.00	-11.66* -48.64* -21.70	-2.54* -69.85* 0.00	±28.6 ±5.96 0.00	-18.90* -49.40* -21.70	-3.27* -1.83* 0.00
13	0.0-0.5 0.6-1.0 1.1-1.5	-2.19* 1.31 -3.76*	±25.62 ±11.34 ±0.30	-7.95* +2.11 -0.16*	-1.59 1.67 0.00	±15.25 ±6.93 0.00	-3.44 +1.64 -0.20	-1.07 -0.25 -1.00
20	0.0-0.5 0.6-1.0	2.90* -13.69*	±30.64 ±11.65	+12.55* -22.55*	2.08* -5.23*	±27.77 ±19.17	+8.18* -14.18*	-0.75 -2.64*
27	0.0-0.5 0.6-1.0 1.1-1.5	-7.04* 7.02* 3.14*	±24.45 ±24.25 ±0.60	-24.34* +24.07* +0.27*	-2.81* 2.78* 1.56	±12.35 ±12.27 ±0.38	-4.90* +4.82* +0.08	5.02* -5.01* -1.82
34	0.0-0.5 0.6-1.0 1.1-1.5 1.6-2.0	-14.66* 15.09* 0.37 -61.50*	±21.65 ±21.23 ±1.45 ±0.06	-44.91* +45.32 +0.08* -0.49	-11.59* 12.12* -0.52 -61.50*	±17.29 ±16.87 ±1.54 ±0.06	-28.33* +28.93* -0.11 -0.49*	-4.23* -4.27* -0.63 0.00
41	0.0-0.5 0.6-1.0 1.1-1.5	-1.58 3.22* -26.76*	±21.37 ±20.82 ±1.26	-4.79 +9.49* -4.78*	7.11* -4.46* -54.91*	±13.40 ±13.17 ±0.67	+13.48* -8.30* -5.18*	5.12* -5.10* -1.95
48	0.0-0.5 0.6-1.0 1.1-1.5	-9.40* 9.51* 5.59*	±16.89 ±15.81 ±1.51	-22.45* +21.26* +1.19*	-9.09* 9.08* 6.31*	±16.25 ±14.78 ±2.04	-20.89* +18.98* +1.82*	0.47 -0.75 1.76
56	0.0-0.5 0.6-1.0 1.1-1.5 1.6-2.0 >2.0	-25.10* 27.16* 10.29* -0.70 -0.86	±12.91 ±9.06 ±7.65 ±0.60 ±0.38	-45.84* +34.80* +11.15* -0.06 -0.05	-23.07* 22.57* 10.61* 2.58* 4.27*	±11.45 ±8.79 ±5.61 ±0.89 ±0.91	-37.38* +28.08* +8.42* +0.32* +0.55*	3.47* -3.76* -2.03* 2.53* 4.28*
63	0.0-0.5 0.6-1.0 1.1-1.5 1.6-2.0 >2.0	-12.69* 16.16* -12.29* 3.03* -5.27*	±13.30 ±12.49 ±2.51 ±0.70 ±0.83	-23.88* +28.55* -4.36* +0.30* -0.62*	-26.81* 28.61* -6.12* 5.17* 1.35	±9.38 ±9.13 ±2.88 ±1.16 ±1.40	-35.55* +36.94* -2.50* +0.85* +0.27	-5.07* 3.83* 3.45* 2.84* 3.86*

I and II; see: Material and methods. Indexed (*) t test statistics values are statistically significant (p<0.05). Indexed (*) L values differed significantly from zero (p<0.05). 1 Critical value t = |2.01| (df 49; α=0.05). 2 Critical value t = |1.98| (df 98; α=0.05).

Although in both stocking densities the food preferences of reared peled in terms of the zooplankton size were usually identical, the statistical analysis showed some significance of differences in food selection between the cages (except days 13 and 48 of the experiment). These differences occurred for two size groups of zooplankton (0.0-0.5 and 0.6-1.0 mm), most numerous both in the environment and in the larval food composition. In the final two weeks of the experiment, such differences were observed for all length classes of the zooplankton (Table 2). Fish larvae reared in the lower stocking density were characterized by a higher selectivity for food (more extreme values of L index), only the last day of the experiment was exceptional. The fish reared in variant II reached larger lengths (TL=64.6 mm) than those from variant I (TL=62.3 mm; Fig. 3).

Peled larvae reared in Lake Legińskie showed some clear differences in feeding preferences between the two stocking density variants (Table 3). The fish from the cage stocked with 10 000 larvae more often avoided small zooplankton, up to 0.5 mm in length. From day 33 onwards, they statistically significantly preferred larger prey, especially those from the 0.6-1.0 mm class, and as the fish grew larger they tended to catch zooplankton from the other size classes. As regards the stocking density of 40 000 larvae per cage, the smallest zooplankton (up to 0.5 mm in length) were much more often under feeding pressure. For larger organisms selection was negative. This was due to inferior food availability in the cage with the higher stocking density. In variant I there were far fewer fish because of the smaller initial stocking volume and higher mortality in the early days of rearing. The fish grew on average to TL=43.6 mm versus TL=30.2 mm in variant II (Fig. 3).

Table 3. The mean values of Strauss food selection index (L) of Coregonus peled larvae reared in illuminated cages in two stocking variants in Lake Legińskie and the results of comparisons of L values between stocking variants

Days of rearing	Plankton length classes (mm)	I			II			Comparison of I and II
		t1	SD	L	t1	SD	L	t2
5	0.0-0.5 0.6-1.0	-46.15* -1.85	±14.14 ±14.14	-92.30* -3.70	0.00 0.00	0.00 0.00	-94.30 -5.70	-1.00 -1.00
12	0.0-0.5 0.6-1.0	-8.82* -0.59	±44.47 ±30.02	-55.50* -2.50	-17.64* 0.00	±30.30 0.00	-75.60* -14.40	-2.64* -2.80*
19	0.0-0.5 0.6-1.0	-2.06* -1.87	±30.19 ±4.46	-8.82* -1.18	-1.75 0.94	±21.09 ±9.12	-5.22 +1.22	0.69 1.67
26	0.0-0.5 0.6-1.0 1.1-1.5	0.55 -1.86 1.32	±18.72 ±13.64 ±0.64	+1.46 -3.58 +0.12	-2.10* 1.70 1.00	±17.31 ±12.93 ±0.16	-5.13* +3.11 +0.02	1.83 2.52* -1.05
33	0.0-0.5 0.6-1.0 1.1-1.5 1.6-2.0	-8.21* 8.25* 4.21* -1.27	±24.47 ±18.42 ±11.75 ±0.23	-28.42* +21.48* +6.99* -0.04	-5.49* 5.62* -5.33* 0.00	±20.64 ±20.61 ±0.34 0.00	-16.03* +16.38* -0.28* -0.10	2.74* -1.30 -4.38* -1.80
40	0.0-0.5 0.6-1.0 1.1-1.5	-15.22* 13.96* 2.51*	±18.61 ±18.23 ±11.19	-40.06* +36.00* +3.97*	2.15* -2.06* -1.89	±18.04 ±17.74 ±1.24	+5.50* -5.16* -0.33	12.43* -11.44* -2.70*
47	0.0-0.5 0.6-1.0 1.1-1.5 1.6-2.0 >2.0	-3.59* 3.55* 5.65* -3.02* -16.78*	±15.56 ±13.74 ±3.29 ±0.65 ±0.56	-7.90* +6.89* +2.63* -0.28* -1.34*	2.17* -7.09* 1.69 0.10 0.50	±28.67 ±15.95 ±1.31 ±4.71 ±11.67	+8.78* -16.00* +0.31 +0.07 +0.83	3.62* -7.69* -4.62* 0.51 1.31
54	0.0-0.5 0.6-1.0 1.1-1.5 1.6-2.0 >2.0	-8.50* 7.59* 5.01* 4.08* 3.54*	±17.80 ±16.42 ±2.59 ±0.78 ±2.91	-21.36* +17.62* +1.84* +0.42* +1.46*	2.94* -3.32* 1.00 1.39 0.19	±9.07 ±8.96 ±0.16 ±0.55 ±1,77	+4.03* -4.21* +0.02 +0.11 +0.05	8.87* -8.25* -4.94* -2.55* -2.92*
61	0.0-0.5 0.6-1.0 1.6-2.0 >2.0	-15.25* 14.80* 2.98* 0.00	±16.18 ±16.18 ±1.65 0.00	-34.89* +33.86* +0.70* -0.10	-3.04* 3.82* 1.26 0.00	±16.59 ±9.02 ±1.46 0.00	-7.12* +4.87* +0.35 -0.10	8.47* -11.06* -1.64 -0.96
68	0.0-0.5 0.6-1.0 1.6-2.0 >2.0	-1.00 0.66 1.75 0.00	±18.76 ±18.41 ±4.44 0.00	-2.66 +1.72 +1.10 -0.30	5.65* -9.61* 1.00 -1.54	±18.16 ±12.04 ±0.21 ±0.73	+14.50* -16.37* +0.03 -0.16	4.65* -5.82* -1.70 1.35

I and II; see: Material and methods. Indexed (*) t test statistics values are statistically significant (p<0.05). Indexed (*) L values differed significantly from zero (p<0.05). 1 Critical value t = |2.01| (df 49; α=0.05). 2 Critical value t = |1.98| (df 98; α=0.05).

There were also noted considerable individual differences between larvae during their foraging activity. For example, one individual sampled on day 33 of rearing from Lake Legińskie (variant I) showed positive selectivity for the 0.0-0.5 mm class at L= +26.7, but its preference for the 1.0-1.5 mm class was close to zero (L= -0.4), whereas the respective preferences of another individual were at the levels of L= -69.6 and L= +48.4.

DISCUSSION

Successful rearing of coregonid fish in illuminated cages depends on environmental conditions, of which the most important ones are temperature [6] and zooplankton abundance [7]. Fluctuations in the cage zooplankton density and composition affect food selection by larval peled. At the beginning of the experiment the zooplankton density corresponded to the thermal conditions in each water body [14]. The lowest concentration of planktonic organisms was found in Lake Legińskie (133 indiv. l^-1, mainly rotifers). Considering the fact that the optimum number of zooplankton is estimated at 200-260 indiv. l^-1, provided the weight of plankton is 0.004 mg indiv.^-1 [4], the plankton concentration in Legińskie Lake could have been insufficient for the peled larvae.

The use of planktonic organisms by fish larvae is determined both by species composition and their size structure [11]. Mouth size and force of suction, related to the fish size, are the parameters which restrict possible dimensions of prey organisms, especially in the first days of fish life [10, 18, 19]. Limitations in food accessibility to coregonid fish first of all concern larvae to 20 mm in length [13]. Larval Coregonus clupeaformis (Mitchill) about 14 mm long did not require prey organisms shorter than 0.5 mm, but had some difficulty consuming copepods up to 2.0 mm long [23]. Vendace (Coregonus albula L.) about 20-30 mm in length consumed specimens of Bosmina coregoni (Baird.) same size as adult fish did, whereas big prey like Bythotrephes sp. and Leptodora kindti appeared in their diet after the fish have attained the length of 40-60 mm [11]. At the beginning of this study reared peled larvae selected zooplankton from the 0.0-0.5 mm length class only, which was caused by the above mentioned limitations in food accessibility. Since at that time the zooplankton in the pond and in the lakes was composed of considerable numbers of larger organisms (longer than 0.5 mm; Fig. 2), larvae found it difficult to catch suitable prey, which was demonstrated by the fact that many of the fish sampled had empty alimentary tract. However, as the larvae grew, they exerted an increasingly stronger food pressure on bigger planktonic organisms. The fish statistically significantly selected or avoided zooplankton to 1.5 mm and fed rather randomly on bigger organisms. Under good foraging conditions (e.g. in Lake Mutek) the reared fish larvae preferred medium-sized prey (classes 0.6-1.0 mm and 1.0-1.5 mm) but avoided the smallest organisms, comprising large numbers of rotifers, particularly Keratella sp. and Kellikotia sp., which are avoided by coregonid larvae [22, 23]. Stronger food pressure on minute organisms appeared at a time of poor feeding conditions, when numbers of zooplankton in the environment declined and the ration of the smallest zooplankton (0.0-0.5 mm) increased. The results obtained from the present study are congruent with the optimal foraging theory (OFT), which states that food choice is based on selecting the most profitable prey while reducing foraging energy cost by choosing the most abundant food organisms [11, 25].

Considerable individual differences observed between larvae during their foraging activity were similar to the results reported by Teska and Behmer [23]. The differences in feeding preferences for organisms of specific lengths were related to the body size of the fish. In the course of the experiment, as the larvae continued to grow, differences in their lengths increased (Fig. 3; bars of the standard deviation). Another study carried out on the fish larvae and zooplankton Lake Legińskie proved that these differences had some effect on the energy value of the ingested zooplankton [8]. Generally, larger fish, which occupied higher positions in the hierarchy of a shoal, consumed food of higher energy value (that is larger prey); the smallest fish had the smallest share in the total energy value of the zooplankton consumed (they had to enjoy smaller prey).

Statistically significant differences in food selection that were observed between larval peled from the cages with different stocking densities concerned mainly smaller zooplankton (0.0-0.5 and 0.6-1.0), which were also the most abundant in the environment and the most willingly caught by the fish larvae. Such differences were recorded even when the feeding preferences of the fish larvae from both stocking density variants were similar (as seen in Lake Mutek and the pond). The fish reared at the lower density, which meant better feeding conditions, would typically prefer larger prey (higher L index values for larger organisms). Under more difficult feeding conditions, found in the cages with the higher stocking density, fish were forced to eat even the smallest zooplankton, which was demonstrated by the less extreme values of Strauss linear index of food selection for these classes of planktonic food in Lake Mutek and in the pond or by the diverse foraging preferences in the both stocking density variants in Lake Legińskie.

CONCLUSIONS

1. At the beginning of rearing peled larvae selected zooplankton up to 0.5 mm in length.

2. As the larvae grew, their preferences changed. Under good foraging conditions larvae statistically significantly selected medium-sized prey (0.6-1.5 mm in length) and avoided the smallest organisms. They fed rather randomly on larger planktonic organisms (>1.5 mm), which were not abundant in the environment.

3. There were observed statistically significant differences in food selection between the cages with different stocking densities, even when the feeding preferences were similar. The fish reared at the lower density would typically prefer larger prey.

REFERENCES

1. Braum E., 1967. The survival of fish larvae in reference of their feeding behaviour and the food supply. In: Biological Basis of Freshwater Fish Production, Oxford, Blackwell Sci. Publ., 113-131.

2. Buskey E. J., Coulter C., Strom S., 1993. Locomotory patterns of microzooplankton: potential effects on food selectivity of larval fish. Bull. of Mar. Sci. 53, 29-43.

3. Champigneulle A., Rojas Beltran R., 1990. First attempts to optimize the mass rearing of whitefish (Coregonus lavaretus L.) larvae from Leman and Bourget Lakes (France) in tanks and cages. Aquat. Liv. Res. 3, 217-228.

4. Dąbrowski K., 1976. How to calculate the optimal density of food for fish larvae. Env. Biol. Fish. 1, 87-89.

5. Drost M. R., Osse J. W. M., Muller M., 1988. Prey capture by fish larvae, water flow patterns and the effect of escape movements of prey. Neth. J. Zool. 38 (1), 23-45.

6. Eckmann R., Pusch M., 1989. The influence of temperature on growth of young coregonids (Coregonus lavaretus L.) in a large prealpine lake. Rapp. P. V. Reun. Cons. Int. Explor. Mer. 191, 201-208.

7. Flüchter J., 1980. Review of the present knowledge of rearing whitefish (Coregonidae) larvae. Aquaculture 19, 191-208.

8. Furgała-Selezniow G., Mamcarz A., 2004. Food resource partitioning by peled larvae (Coregonus peled Gmel.) of various size, reared in illuminated cages at different densities. Acta Sci. Pol., Piscaria 3 (1-2), 23-32.

9. Galliot M. F., Maufoy C., 1983. Elevage larvaire de coregones en cages, bacs, etangs. Mem. Ecol. Sup. Techn., Biol. Appl., 53.

10. Hartmann J., 1986. Interspecific predictors of selected prey of young fishes. Arch. Hydrobiol. 22, 373-386.

11. Huusko A., Sutela T., Karjalainen J., Auvinen H., Alasaarela E., 1988. Feeding of vendace (Coregonus albula L.) fry in a natural-state lake and a regulated lake in Northern Finland. Finn. Fish. Res. 9, 447-456.

12. Luczynski M., Falkowski S., Kopecki T., 1988. Larval development in four coregonid species (Coregonus albula, C. lavaretus, C. muksun and C. peled). Finn. Fish. Res. 9, 61-69.

13. Mamcarz A., 1990a. Uwarunkowania wzrostu larw pelugi (Coregonus peled Gmel.) w czasie podchowu sadzowego [Conditions for growth of Coregonus peled larvae in cage rearing]. Acta Acad. Agric. Tech. Olst., Prot. Aqua. Piscat. 17, 1-57 [in Polish].

14. Mamcarz A., 1990b. Studies on cage rearing of the coregonid fishes (Coregonidae) in different water bodies. I. Characteristics of abiotic factors. Acta Acad. Agric. Tech. Olst., Prot. Aqua. Piscat. 18, 47-64.

15. Mamcarz, A., 1993. Effect of altered prey density on feeding and growth of Coregonus albula L. larvae (in: Physiological and Biochemical aspects of Fish larval Development) Eds. B.T. Walther, H.J. Fyhn. Univ. of Bergen, Norway, 172-179.

16. Mamcarz A., Nowak M., 1987. New version of an illuminated cage for coregonid rearing. Aquaculture 65, 183-188.

17. Okubo A., 1972. A note on small organism diffusion around an attractive center: a mathematical model. J. Oceanogr. Soc. Jap. 28, 1-7.

18. Ponton D., Müller R., 1988. Distribution and food of larval and juvenile Coregonus sp. in Lake Sarnen, Switzerland. Finn. Fish. Res. 9, 117-125.

19. Ponton D., Müller R., 1990. Size of prey ingested by whitefish, Coregonus sp., larvae. Are Coregonus larvae gape-limited predators? J. Fish Biol. 36, 67-72.

20. Sandlund O. T., Naesje T. F., Kjellberg G., 1987. The size selection of Bosmina longispina and Daphnia galatea by co-occuring cisco (Coregonus albula), whitefish (C. lavaretus) and smelt (Osmerus eperlanus). Arch. Hydrobiol. 110, 357-363.

21. Strauss R. E., 1979. Reliability estimates for Ivlev´s electivity index, the forage ratio, and a proposed linear index of food selection. Trans. Amer. Fish. Soc. 108, 344-352.

22. Sutela T., 1998. Feeding of coregonid larvae and digestive degradation of prey zooplankton. Applications of a gut segmentation method. Acta Univ. Oulu. Ser. A. Sci. Rerum Nat. 314, 1-29.

23. Teska J. D., Behmer D. J., 1981, Zooplankton preferences of larval lake whitefish. Trans. Am. Fish. Soc. 110, 459-461.

24. Weihs D., 1980. Energetics significiance of changes swimming modes during growth of larval anchovy, Engraulis modrax. Fish. Bull. U.S. 77, 597-604.

25. Werner E. E., Hall D. J., 1974. Optimal foraging and the size selection of prey by the bluegill sunfish (Lepomis macrochirus). Ecology 55, 1042-1052.

---

Grażyna Furgała-Selezniow
Department of Lake and River Fisheries,
University of Warmia and Mazury in Olsztyn, Poland
M. Oczapowskiego 5, 10-957 Olsztyn, Poland
phone +48 89 523 33 88,
fax +48 89 523 39 69
email: graszka@uwm.edu.pl

Andrzej Mamcarz
Department of Lake and River Fisheries,
University of Warmia and Mazury in Olsztyn, Poland
M. Oczapowskiego 5, 10-957 Olsztyn, Poland
phone +48 89 523 33 88,
fax +48 89 523 39 69
email: mamcarz@uwm.edu.pl

Andrzej Skrzypczak
Department of Lake and River Fisheries,
University of Warmia and Mazury in Olsztyn, Poland
M. Oczapowskiego 5, 10-957 Olsztyn, Poland
phone +48 89 523 33 88,
fax +48 89 523 39 69
email: sandacz@uwm.edu.pl

---

Responses to this article, comments are invited and should be submitted within three months of the publication of the article. If accepted for publication, they will be published in the chapter headed 'Discussions' and hyperlinked to the article.

---

Main  -  Issues  -  How to Submit  -  From the Publisher  -  Search  -  Subscription