- Email: [email protected]
A comparison of methods for calculating Catch Per Unit Effort (CPUE) of gill net catches in lakes
Fisheries Research 93 (2008) 204–211
Contents lists available at ScienceDirect
Fisheries Research journal homepage: www.elsevier.com/locate/fishres
A comparison of methods for calculating Catch Per Unit Effort (CPUE) of gill net catches in lakes Torben L. Lauridsen a,∗ , Frank Landkildehus a , Erik Jeppesen a,b , Torben B. Jørgensen a , Martin Søndergaard a a National Environmental Research Institute, University of Aarhus, Department of Freshwater Ecology, Vejlsøvej 25, P.O. Box 314, 8600 Silkeborg, Denmark b University of Aarhus, Department of Biological Sciences, Plant Biology, Ole Worms All´e, Building 135, DK-8000 Aarhus C, Denmark
a r t i c l e
i n f o
Article history: Received 7 February 2008 Received in revised form 22 April 2008 Accepted 22 April 2008 Keywords: Fish Gill nets CEN standard Pelagic Benthic Water Framework Directive
a b s t r a c t Fish constitute an important component of lake ecosystems and many different methods have been used for fish assessment. Based on gill net catches in two stratified (max depth = 14–22 m) eutrophic Danish lakes, relative fish abundance measured as Catch Per Unit Effort (CPUE) was calculated. We used three different methods of which two followed the European standard based on benthic nets (CEN (European Committee for Standardization), 2005: EN 14757. Water quality – Sampling of fish with multi-mesh gill nets. Brussels, 27 pp.), one assuming equal volumes in all depth strata and the other using calculated volumes in the depth strata. The third method followed a modified CEN standard, adopted as a new Danish (DK) standardized method based on calculated benthic and pelagic water volumes and by including both benthic nets and, compared to the CEN standard, an increased fishing effort with pelagic nets. Fish were concentrated in the littoral/benthic part of the upper two depth strata (0–6 m depth) with an up to 8 fold higher abundance than in the pelagic. Calculated CPUE is highly sensitive to the morphometry of the lakes. In lakes with extreme morphometry (unequal volumes in depth strata) it is important to use calculated water volumes for the depth strata. By including information derived from the pelagic nets, total lake CPUEs were 42–56% lower than CPUE values based on benthic nets only. We further show that the relative contribution of CPUE between habitats changes markedly with the nutrient level in 12 deep lakes. It is concluded that for deep lakes it is of key importance to include pelagic nets when comparing fish assemblages and abundances among lakes and when evaluating effects of major changes in key environmental factors, such as nutrient loading and climate. © 2008 Elsevier B.V. All rights reserved.
1. Introduction It is generally accepted that fish strongly influence trophic dynamics in lakes, occasionally with great impact on phytoplankton and on the water quality and ecological state in both freshwater lakes (Carpenter and Kitchell, 1993; Rudstam et al., 1993; Jeppesen et al., 1997, 2000) and brackish lakes (Jakobsen et al., 2004). Fish also contribute to biodiversity and are of interest to both anglers and commercial fishermen. Moreover, fish are one of the four biological variables to be monitored according to the EU Water Framework Directive (WFD) for European water bodies (European Union, 2000). Knowledge of fish community structure and abundance is therefore of key importance. Various gears have been used
∗ Corresponding author. Tel.: +45 89 20 14 00; fax: +45 89 20 14 14. E-mail address: [email protected] (T.L. Lauridsen). 0165-7836/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.fishres.2008.04.007
for fish surveys, such as seine nets, fyke nets, trawls, gill nets, electrofishing and hydroacoustics (e.g., Mehner et al., 2005a; Prchalova´ et al., 2006; Juza and Kubeˇcka, 2007), or assessments have been based on commercial catch data. In general, the methods follow traditions, laws and regulations and the specific sites and aims. However, the use of different gear renders comparisons of fish-data between lakes difficult (Lauridsen et al., 1999; Appelberg, 2000), and even if similar gear has been used, dimensions will most likely differ (Mehner et al., 2007). Results from the marine environment have indicated that an understanding of gear dynamics and environmental influences as well as of habitat and depth influence is important for analyzing CPUE (Bigelow and Maunder, 2007). With the implementation of the EU WFD, it is claimed to use general standards in lake monitoring. Such a standard has not been adopted for fish so far; however, a European standard on gill netting (CEN, 2005) was introduced in 2005. To obtain a standardized Catch Per Unit Effort (CPUE) for fish weight and numbers,
T.L. Lauridsen et al. / Fisheries Research 93 (2008) 204–211
205
Fig. 1. Map of Denmark showing the two study lakes, Lake Brassø and Lake Mossø, in the River Gudena˚ system. Hypsographs illustrate the extreme morphomotry of Lake Brassø and the normal morphometry of Lake Mossø, see text for further explanation.
guidelines for field sampling using a range of depth strata and benthic gill nets combined with pelagic nets at a single site in each lake have been established. In Denmark, a standardized gill netting method (Mortensen et al., 1990) was introduced in 1989 and used in approximately 120 lakes until 2004. The method by Mortensen et al. (1990) prescribed use of benthic, pelagic and floating Lundgreen gill nets (length 42 m, height 1.5 m, multiple mesh with fourteen 3-m panels ranging from 6.25 to 75 mm) together with electrofishing in the littoral zone. The Mortensen et al. (1990) method was not, however, directly comparable to the Swedish and Finnish standards using Nordic gill nets (Appelberg, 2000), which later became the CEN standard (CEN, 2005). To ensure comparability of data, it was decided to substitute the Mortensen et al. (1990) method with the CEN standard (CEN, 2005) for surveys in Danish lakes. However, to increase the effort on pelagic fishes, randomly distributed pelagic gill nets were introduced into the Danish (DK) version of the CEN standard, as pelagic catches may vary considerably from benthic catches and the relative contribution may change with eutrophication (Jeppesen et al., 2006). Such changes may not be registered by just using benthic nets and the change in relative abundance using the CEN standard
Table 1 Characteristics of the two study lakes
Area (ha) Maximum depth (m) Mean depth (m) Retention time (yr) Total phosphorus (g l−1 )
Lake Brassø
Lake Mossø
122 14 4.6 0.01 101 (2005)
1690 22 9 2.2 93 (2005)
is actually not comparable. The DK method allows calculations of CPUE in different ways, of which one is comparable to the CEN standard and the other takes into account the pelagic gill nets. The aim of this study is to demonstrate how the inclusion of pelagic gill nets and the diversion of depth strata into benthic and pelagic zones affect the calculated total fish CPUE in deep lakes. A second aim is to evaluate the necessity of using similar methods in the CPUE data processing.
2. Study area Lake Brassø (122 ha, max. depth 14 m) and Lake Mossø (1690 ha, max depth 22 m) are both situated in the central part of Jutland (Denmark) (Fig. 1). They are part of the River Gudena˚ system and by Danish standard; they are large, eutrophic and deep lakes with low Secchi depth and sparse submerged vegetation. For more morphometric and abiotic characteristics, see Table 1. Since the lakes are large and situated geographically closely in the same river system, their fish stock compositions are very similar (Table 2). Lake Brassø has a deeper part in its eastern end in which the deepest spot covers only a minor area (Fig. 1); a characteristic defined as ‘extreme morphometry’ according to the CEN standard. Lake Mossø has a large deep part in the eastern end but no small deep spot, and its morphemetry is therefore ‘normal’ according to the CEN standard (Fig. 1). In the period June to August and during stratification oxygen concentrations are below 6 mg O2 l−1 in the bottom water (>18 m) of Lake Mossø (Jørgensen, 2000). Lake Brassø is not permanently stratified during summer; however, temporary oxygen depletion does occur in the bottom water (>6 m, T. Jørgensen, unpubl. results).
206
T.L. Lauridsen et al. / Fisheries Research 93 (2008) 204–211
Table 2 List of fish species caught in Lake Mossø and Lake Brassø
Bream, Abramis brama Eel, Anguilla anguillla Gudgeon, Gobio gobio Whitefish, Coregonus spp. Pike, Esox lucius Pikeperch, Sander lucioperca Perch, Perca fluviatilis Roach, Rutilus rutilus Ruffe, Scardinius erythropthalmus Smelt, Osmerus eperlanus Three-spined stickleback, Gasterosteus aculeatus Nine-spined stickleback, Pungitius pungitius
Lake Brassø
Lake Mossø
X X
X X X X X X X X X X X
X X X X X X X X X
3. Material and methods Fishing was performed with Nordic Norm gill nets between August 15 and September 15 according to the DK standard (Lauridsen et al., 2005). According to the CEN standard additional meshes can be added if large fish of certain species are hard to catch (CEN, 2005). Thus, due to the presence of large bream (Abramis brama) in Danish lakes, two extra mesh sizes (68 and 85 mm) were added at the end of the 43 mm mesh of the CEN standard nets, yielding a total of 14 different mesh sizes ranging from 5 to 85 mm knot to knot following a 1.25 geometric series and a total length of 35 m. The sampling procedure is based on stratified random sampling (Appelberg, 2000). The sampled lakes are divided into depth
strata and random sampling is performed within each stratum. By randomizing each gill net within each depth stratum and by randomizing the angle an independent sample was obtained (CEN, 2005). Due to the general shallowness of Danish lakes the DK standard operates with 4 depth strata: 0–3, 3–6, 6–12 and >12 m (Table 3). As both lakes are deeper than 10 m, pelagic nets were added as recommended in the CEN standard. However, following the DK standard, the additional nets are used to undertake an additional stratified random sampling in the pelagic part of each depth stratum, resulting in a minimum of 3 nets in both the pelagic and benthic parts of each stratum (Fig. 2). In the two studied lakes 3–4 benthic and pelagic nets, respectively, were used per stratum. The nets are 1.5 m deep. A standardized fishing period lasted 14–16 h and the nets were set between 4 and 6 p.m. and lifted the following morning between 6 and 8 a.m. Catch Per Unit Effort can be calculated for the total fish stock and for each species. CPUE is defined as a water volume weighted average catch in one net, calculated in three different ways: (1) as whole lake CPUE based on the CEN standard (CEN, 2005) using benthic gill nets and either assuming equal volume in the different depth strata (CEN eqlvol) or (2) using actual volumes in the depth strata (CEN calcvol), or (3) as whole lake CPUE based on the DK standard (Lauridsen et al., 2005) using both benthic and pelagic gill nets and actual volumes in the different parts of the depth strata: CPUEDK =
∗ ∗ ∗ ∗ CPUEben1 + Vpel1 CPUEpel1 + . . . .. + Vben4 CPUEben4 + Vpel4 CPUEpel4 ) (Vben1
Vtotal
Vben1 is the volume of the near-shore area in depth stratum 1; the area between the 0- and the 3-m depth curve multiplied with 1.5
Table 3 Sampling effort in lakes according to the DK (Danish) method (Lauridsen et al., 2005) Maximum depth (m) Area (ha)
Depth strata (m)
−4.5 (# ben)
>4.5–7.5 (# ben/# pel)
>7.5–13.5 (# ben/# pel)
>13.5 (# ben/# pel)
3/3 3
3/3 3/3 x/
3/3 3/3 x/x x 12
< 20
−3* >3–6 >6–12 >12 Tot net
6
6
9
12
20–50
−3* >3–6 >6–12 >12 Tot net
6
3/3 3
3/3 3/3 3
6
9
15
51–100
−3* >3–6 >6–12 >12 Tot net
8
4/4 4
3/3 3/3 3/x
8
12
15
101–250
−3* >3–6 >6–12 >12 Tot net
10
4/4 4
3/3 3/3 3/x
10
12
15
251–1000
−3* >3–6 >6–12 >12 Tot net
12
4/4 4
4/3 4/3 4/x
12
12
18
> 1000
−3* >3–6 >6–12 >12 Tot net
14
6/6 6
4/3 4/3 4/x
18
18
14
3/3 3/3 3/x x/ 15 3/3 3/3 3/x x/ 15 3/3 3/3 3/3 3/x 21 4/3 3/3 3/3 3/x 22 4/3 3/3 3/3 3/x 22
The table shows numbers (#) of nets and how benthic and pelagic nets shall be distributed in the different depth strata at given areas (first column) and given maximum depths (second row). ben, benthic nets; pel, pelagic nets; x, indicates where to place additional nets if maximum lake depth >10 m; (*) if maximum lake depth is ≤4.5 m, depth stratum 1 is expanded to the maximum depth.
T.L. Lauridsen et al. / Fisheries Research 93 (2008) 204–211
Fig. 2. (A) An example of the distribution of benthic and pelagic gill nets in a lake with several depth strata. According to the DK standard (Danish standard) at least 3 nets are set in both the benthic and pelagic part of each depth stratum and (B) division of depth strata volumes into benthic and pelagic parts.
(0.5 × the difference between lower and upper threshold for the depth stratum), and Vpel1 is the corresponding volume in the pelagic part of depth stratum 1; (the area with a depth >3 m) × (the difference between lower and upper threshold for the depth stratum) (Fig. 2). Vtotal is the total lake volume. CPUE can be calculated with and without the 68 mm and 85 mm meshes. In the present study CPUE is calculated without the extra meshes as NPUE (numbers) and WPUE (weight), this being the most frequent standard. The calculations of the depth strata volumes are based on the hypsographs in combination with the bathymetric maps of the lakes (Fig. 1). The latter is used to calculate the areas within the different depth strata. In literature, the percentage of piscivores is suggested as a useful indicator of water quality (Søndergaard et al., 2005). Consequently, the potential piscivore biomass, including all size classes of perch, pike and pikeperch (Stizostedion lucioperca L), measured as percentage (%) of the total fish biomass, is calculated using the three different methods mentioned above. To elucidate how the relative contribution of benthic and pelagic catches changes with the nutrient level, we also compared data collected from 12 deeper lakes (mean depth larger than 6 m), in which only the previously used Mortensen et al. (1990) measure and Lundgreen gill nets were used. Average mean depth of these 12 lakes was 10.2 m and the lake areas ranged between 27 and 1730 ha (Table 4). Since the Mortensen et al. (1990) method did not divide lakes into depth strata (for details see Jeppesen et al., 2006) we calculated average WPUEs for littoral benthic nets and float-
Table 4 Total summer mean phosphorus concentration (TP) and morphometric data on the 12 lakes in which gill netting according to Mortensen et al. (1990) was performed Lake no.
TP (g l−1 )
Mean depth (m)
Max. depth (m)
Area (ha)
1 2 3 4 5 6 7 8 9 10 11 12
114 134 23 135 210 41 210 56 109 32 209 146
10.3 14.6 15 7.1 8 8.7 13.5 13.5 6.3 6.7 9.1 9.9
22 34 33 16 19 15
1301 332 180 315 370 27 1730 941 120 76 1250 662
38 15 14
207
ing nets, respectively, situated 1/2 of the distance from the shore to the lake middle. These results are defined as littoral and pelagic surface results (depth <3 m). Representing the benthic and pelagic of the deeper parts of the lakes we calculated average WPUEs for the benthic nets situated 3/4 of the distance from the shore to the lake middle and for the pelagic nets situated in the middle of the water column and 1/2 of the distance from the shore and the lake middle. These results are defined as benthic and pelagic meta-/hypolimnic samples (depth >6 m). From a pelagial station a depth integrated water sample was collected fortnightly during the period May to October from the epilimnion for total phosphorus (TP). TP was determined as molybdate reactive phosphorus (Murphy and Riley, 1962) following persulphate digestion (Koroleff, 1970). In each lake, habitat effects on total CPUE were tested using a one-way ANOVA on log-transformed fish-data. When the habitat effect was significant, Tukey’s test was applied to data to test for differences between the habitats, between the depth strata and within the depth strata (benthic vs. pelagic nets), at a 0.05 significance level. Similar statistics were applied on the data for the four dominant fish species in the lakes. In the 12 additional lakes, sampled according to Mortensen et al. (1990), we tested for general differences between the littoral and pelagic surface habitats, using a paired t-test on average WPUE values for the two habitats in the 12 lakes. In addition, linear Pearson correlations were performed on benthic littoral, benthic meta-/hypolimnic and pelagic epilimnic WPUE data, respectively, vs. total phosphorus. For unimodal relations, a second order polynomial regression was performed.
4. Results In Lake Brassø there was a significant habitat effect on CPUE (ANOVA test on log-transformed data, p = 0.049 and p = 0.0094 for WPUE and NPUE, respectively), CPUE being highest in the upper depth strata compared to the lowest depth strata. When comparing the benthic catches, WPUE was significantly higher in depth stratum 1 than in 4, and NPUE was significantly higher both in depth strata 1 and 2 than in depth stratum 4 (Tukey’s test, ˛ = 0.05) (Fig. 3A). Within the depth strata, CPUE was higher in the littoral benthic part than in the pelagic part of depth strata 1 and 2 (Fig. 3A); however, the effect was not significant on basis of neither weight nor number (Tukey’s test). In depth strata 3 and 4, CPUE were low and similar between the benthic and pelagic parts (Fig. 3A). In Lake Brassø, 56% of the water volume is in the 1st depth stratum (0–3 m), 29% in the 2nd depth stratum 2 (3–6 m), 14% in the 3rd depth stratum (6–12 m) and 0.5% in the 4th depth stratum (>12 m). Using the CEN eqlvol method, total WPUE and NPUE were 4.51 ± 1.24 (avg ± std) kg net−1 and 146 ± 47 indiv net−1 , respectively (Fig. 3B). For CEN calcvol, WPUE and NPUE were 6.00 ± 1.67 kg net−1 and 200 ± 68 indiv net−1 , respectively (Fig. 3B), and for DK (Fig. 3B) WPUE and NPUE were 2.54 ± 0.63 kg net−1 and 84 ± 22 indiv net−1 (Fig. 3B), respectively. The percentage of the potential piscivore biomass varied between 27 and 32% of the total WPUE (Fig. 3B). Lake Mossø is a lake with “normal” morphometry. Thirty percent of the total water volume is in the 1st depth stratum, and 22, 28 and 20% in the 2nd, 3rd and 4th depth strata, respectively. Like in Lake Brassø, there was a significant habitat effect on CPUE (ANOVA test on log-transformed data, p = 0.035 and p = 0.0006 for WPUE and NPUE, respectively), with CPUE being higher in the littoral benthic part than in the pelagic part in depth strata 1 and 2. However, only in depth stratum 2 did NPUE show a significant difference, and the remaining comparisons were insignificant, both
208
T.L. Lauridsen et al. / Fisheries Research 93 (2008) 204–211
Fig. 3. (A) Weight in g per net (WPUE ± std) and numbers (NPUE ± std) per unit effort in the benthic and pelagic parts of the 4 depth strata in Lake Brassø and (B) average WPUE ± std and NPUE ± std, based on the CEN standard (CEN, 2005), assuming similar water volumes in the depth strata (CEN eqlvol), the CEN standard using calculated water volumes in the depth strata (CEN calcvol) and the DK standard using calculated water volumes and benthic and pelagic gill nets (DK). Percentage of potential piscivore WPUE (perch, pike and pikeperch) is given in numbers.
on weight and number basis (Tukey’s test, ˛ = 0.05) (Fig. 4A). Comparisons of benthic catches showed no significant differences in WPUE between depth strata 1 and 2 and depth stratum 4, but with regard to NPUE the benthic catches were significantly higher in depth strata 1 and 2 compared to depth stratum 4 (Tukey’s test, ˛ = 0.05) (Fig. 4A). Using the CEN eqlvol method, total WPUE and NPUE were 4.27 ± 1.03 kg net−1 and 121 ± 31 indiv net−1 , respectively (Fig. 4B). CEN calcvol gave WPUE and NPUE values of 4.43 ± 1.21 kg net−1 and 124 ± 35 indiv net−1 , respectively (Fig. 4B), and the DK standard, including benthic and pelagic gill nets and calculated water volumes (DK), yielded WPUE and NPUE values of 1.91 ± 0.77 kg net−1 and 45 ± 24 indiv net−1 , respectively (Fig. 4B). Percentage of the potential piscivore biomass varied between 49 and 55% of the total WPUE (Fig. 4B). In both lakes the dominant fish species were perch (Perca fluviatilis), roach (Rutilus rutilus), ruffe (Scardinius erythropthalmus) and bream. Based on the DK method these four species constituted 89% by weight of the total fish stock in Lake Brassø and 64% in Lake Mossø (Fig. 5). In Lake Mossø, perch and roach were evenly distributed, whereas roach was dominant in Lake Brassø, constituting 40% of the total fish stock on weight basis (Fig. 5). The distributional pattern of the four dominant species was similar to the overall fish distribution (Figs. 3 and 4); thus, there were more fish in the benthic parts than in the pelagic parts of the different depth strata (Fig. 5). Except for NPUE of bream in Lake Brassø and WPUE and NPUE of bream in Lake Mossø, there was a significant habitat effect on CPUE (ANOVA test on log-transformed data). In Lake Brassø, comparisons of benthic catches of perch showed significant differences in WPUE between depth strata 1 and 2 and depth stratum 4. With regard to NPUE, a similar result was found. Additionally, the benthic littoral perch-NPUE was significantly higher than the epilimnic pelagic NPUE (Tukey’s test, ˛ = 0.05)
(Fig. 5A). In Lake Brassø, there was no benthic catches of bream in depth strata 3 and 4 (Fig. 5A). Regarding roach, benthic catches in depth strata 1 and 2 were significantly higher compared to NPUE in depth stratum 4 (Tukey’s test, ˛ = 0.05; Fig. 5A). All epilimnic ruffe of Lake Brassø were caught in the benthic littoral part. In depth strata 2 and 3 ruffe were evenly distributed between the benthic and pelagic parts (Fig. 5A). In Lake Mossø, comparisons of benthic catches of perch showed a significantly higher NPUE in depth strata 1 and 2 than in depth strata 3 and 4 (Tukey’s test, ˛ = 0.05) (Fig. 5B). Comparisons of benthic and pelagic catches showed significantly higher benthic NPUE values in depth strata 1 and 2 (Fig. 5B). Bream were caught in very limited numbers and only in the benthic parts of the four depth strata (Fig. 5B). Roach were not caught in depth stratum 4. Comparisons of benthic roach catches showed significantly higher WPUE and NPUE in depth strata 1 and 2 than in depth strata 3 and 4. Due to high variability between replicates no significant differences were observed between benthic and pelagic catches (Fig. 5B). Although the number of ruffe caught was high, the biomass was relatively low. Regarding both WPUE and NPUE, significantly higher abundances were caught in the benthic parts than in the pelagic parts (Tukey’s test, ˛ = 0.05, Fig. 5B). In the 12 lakes sampled according to the previous Danish protocol (Mortensen et al., 1990), littoral WPUE values were in general higher than the pelagic surface WPUE values (p < 0.0001, paired t-test). Additionally, there was a positive correlation between littoral WPUE and total phosphorus, ranging from 3 to 8 kg net−1 in the low-nutrient lakes to 6–13 kg net−1 in the high-nutrient lakes (Fig. 6A). In the meta-/hypolimnic benthic part there was a negative correlation between WPUE and total phosphorus, ranging from 2 to 7 kg net−1 on average in the low-nutrient lakes to 1–5 kg net−1 in the high-nutrient lakes (Fig. 6A). In the corresponding meta/hypolimnic pelagic samples a unimodal relationship was observed, with the highest catch at nutrient levels around 0.1–0.2 mg TP l−1 (Fig. 6B), while lower WPUE values were observed in the pelagic surface water, irrespective the nutrient level (Fig. 6B).
Fig. 4. (A) Weight in g per net (WPUE ± std) and numbers (NPUE ± std) per unit effort in the benthic and pelagic parts of depth strata in Lake Mossø and (B) average WPUE ± std and NPUE ± std. For further explanations, see Fig. 3. Percentage of potential piscivore WPUE is given in numbers.
T.L. Lauridsen et al. / Fisheries Research 93 (2008) 204–211
209
Fig. 5. The percentual biomass contribution of dominant fish species (perch, roach, bream and ruffe) and others (see Table 3 for species) to the total fish stock, and the distribution of the dominant fish species between pelagic and benthic parts of the four depth strata (mean ± std) in Lake Brassø (A) and Lake Mossø (B). The percentual contribution is calculated according to the DK method (for explanation see Fig. 3). ‘*’ Significant at 0.05 level (Tukey’s test).
5. Discussion In the present study more fish were caught in the littoral, benthic parts than in the pelagic zone in the upper two depth strata of Lake Brassø and Lake Mossø. A similar pattern is seen in several previous studies from deep lakes and reservoirs (e.g., Prchalova´ et al., 2006; Fischer and Eckmann, 1997) and shallow lakes (Lewin et al., 2004; Romare et al., 2003). This distribution pattern was also in accordance with the general pattern observed in the 12 lakes sampled with the previous Danish method (Mortensen et al., 1990). A clear difference in fish abundance between the littoral and the pelagic was found particularly in the more nutrient rich lakes. In the two lower depth strata of Lake Brassø and Lake Mossø fish abundance was generally low irrespective of habitat. If oxygen levels and food availability were sufficient we might have expected to encounter more fish at these depths according to the findings in more oligotrophic systems (Prchalova´ et al., 2006; Rowe et al., 2001; Halvorsen et al., 1997). However, the present pattern can be due to limited oxygen levels (Vasek et al., 2004; Jørgensen, 2000) and less food availability (Gliwicz et al., 2006), affecting the presence of fish (Gliwicz et al., 2006; Vasek et al., 2004). This is supported by data from the 12 additional Danish lakes presented in this study, as the highest benthic catches in the lower depth strata are found at low-nutrient levels and reduced catches occur with increasing total phosphorus and probably reduced oxygen levels. Jeppesen et al. (2006) revealed distributional patterns similar to those in the present study as being typical for deeper eutrophic lakes dominated by cyprinid and percid fish species. Except for bream, the four dominant fish species in the two lakes exhibited the same pattern, with high abundance in the benthic parts of the upper depth strata. Particularly in Lake Mossø, the few bream caught were more evenly distributed between the four depth strata (Fig. 5B). The uneven distribution pattern has marked effects on total average CPUE obtained by the different methods used. If equal water volumes were assumed for all depth strata (CEN eqlvol) and
only benthic nets were used, we overestimated total CPUE, as the high CPUE values from the littoral represent the entire depth strata despite the lower actual fish CPUE in the pelagic parts. Also, use of calculated water volumes for the depth strata in combination with benthic nets, which is in accordance with the CEN standard, leads to overestimation of total fish CPUE. According to the CEN standard (CEN, 2005), Lake Brassø has an ‘extreme morphometry’ with large volume variation between the different depth strata. This was clearly evidenced by our results, since CEN calcvol CPUE (calculated volumes) was 33% higher than the CEN eqlvol CPUE (equal volumes). This is basically due to the large proportion of the total lake volume occurring in depth stratum 1 combined with the large CPUE in the littoral. In contrast, Lake Mossø belongs to the category ‘normal morphometry’. The actual volumes in the depth strata are quite similar and, consequently, there was only minor difference between the CEN calcvol CPUE and the CEN eqlvol CPUE (the latter being 3.7% higher). The results underline the importance of using actual volumes when calculating CPUE in lakes, not least when they belong to the ‘extreme morphometry’ category. In the CEN standard, it is assumed that benthic net catches represent all depth strata, but this seems not to be true for Lake Brassø and Lake Mossø, as overall lower CPUE was obtained in the pelagic than in the benthic parts. Consequently, total CPUE calculated according to the DK method for Lake Brassø was 42–56% lower than when using the official CEN CPUE method, and 55% lower for Lake Mossø. We found that the CPUE based percentage piscivore biomass is almost similar in the two study lakes irrespective of the method used, which is important if this measure is to be used as a water quality indicator (Søndergaard et al., 2005). It can be claimed that the bias introduced by using the official CEN method is similar and that the data from different lakes are thus comparable. However, the data from the additional 12 lakes and an earlier study by Jeppesen et al. (2006) show that the differences between the two methods depend on the nutrient level
210
T.L. Lauridsen et al. / Fisheries Research 93 (2008) 204–211
Acknowledgements We would like to thank the Danish counties and in particular County of Aarhus for access to data, Anne Mette Poulsen for manuscript assistance and Tinna Christensen and Juana Jacobsen for layout assistance. This work was financially supported by the EU EuroLimpacs project (GOCE-CT-2003-505540) and the CLEAR project (a Villum Kann Rasmussen Centre of Excellence Project). References
Fig. 6. Data represent 12 deeper lakes (mean depth >6 m) and encompass 1–3 lake years. Weight per unit effort (WPUE ± std) in (A) benthic nets from the littoral (LitBen, depth <3 m, n = 12–24) and 3/4 of the distance from the shore to the lake middle (HypoBen, depth >6 m, n = 4–12) and (B) floating nets (Epi, n = 5–6) situated 1/2 of the distance from the shore to the lake middle and pelagic meta-/hypolimnic nets (Hypo) situated in the middle of the water column and 1/2 of the distance from the shore to the lake middle (depth >6 m, n = 6–12). Pearson correlation coefficients and a second order polynomial regression coefficient (on the unimodal response) are given.
and are lower in the more oligotrophic systems. Consequently, an evaluation of the effects of changes in nutrient loading cannot be done properly without considering also the pelagic net data. Moreover, species richness and diversity are higher in benthic and littoral habitats compared to pelagic habitats (Diekmann et al., 2005; Mehner et al., 2005b), which suggests that the CEN standard will not give a true picture of the proportion of the different fish species in the lake. Furthermore, the spatial distribution and survival of fish are strongly influenced by the temperature and oxygen content. Vendace (Coregonus albula), for instance, prefer cold hypolimnic water and may also display habitat shifts if conditions change, e.g., the oxygen content (George et al., 2006). Such a habitat shift may wrongly be interpreted as a reduction in vendace CPUE according to the CEN standard. Conditions may also change with alterations in influencing stressors such as climate and nutrients. In Danish lakes, for example, increasing abundances of, for instance, roach and perch occur in the littoral with increasing total phosphorus concentrations (Jeppesen et al., 2006). Consequently, a reduction in nutrient loading may lead to a behavioural change towards the pelagic (Jacobsen et al., 2004; Jeppesen et al., 2006), which could be detected in the CEN standard as reduced CPUE. We therefore emphasize the importance of including pelagic data. Our results are based on relatively few lakes, but underline the importance of using a common method when comparing CPUE values between lakes and regions.
Appelberg, M. (Ed.), 2000. Swedish standard methods for sampling freshwater fish with multi-mesh gillnets. Fiskeriverket information, 1. Bigelow, K.A., Maunder, M.N., 2007. Does habitat or depth influence catch rates of pelagic species? Can. J. Fish. Aquat. Sci. 64, 1581–1594. Carpenter, S.R., Kitchell, J.F. (Eds.), 1993. The Trophic Cascade in Lakes. Cambridge University Press, New York. CEN (European Committee for Standardization), 2005: EN 14757. Water quality – Sampling of fish with multi-mesh gill nets. Brussels, 27 pp. ¨ Diekmann, M., Bramick, U., Lemcke, R., Mehner, T., 2005. Habitat-specific fishing revealed distinct indicator species in German lowland lake fish communities. J. Appl. Ecol. 42, 901–909. European Union, 2000. Directive 2000/60/EC of the European Paliament and of the Council Establishment a Framework of the Community Action in the Field of Water Policy. European Commision, off. J. Eur. Commun. L327, 1. Fischer, P., Eckmann, R., 1997. Spatial distribution of littoral fish species in a large European lake, Lake Constance, Germany. Arch. Hydrobiol. 140, 91–116. George, D.G., Bell, V.A., Parker, J., Moore, R.J., 2006. Using a 1-D mixing model to assess the potential impact of year-to-year changes in weather on the habitat of vendace (Coregonus albula) in Bassenthwaite Lake, Cumbria. Freshw. Biol. 51, 1407–1416. Gliwicz, Z.M., Slon, J., Szynkarczyk, I., 2006. Trading safety for food: evidence from gut contents in roach and bleak captured at different distances offshore from their daytime littoral refuge. Freshw. Biol. 51, 823–839. Halvorsen, M., Jørgensen, L., Amundsen, P.-A., 1997. Habitat utilization of juvenile Atlantic trout (Salmo trutta L.) and Arctic charr (Salvelinus alpinus L.) in two lakes in northern Norway. Ecol. Freshw. Fish 6, 67–77. Jacobsen, L., Berg, S., Jepsen, N., Skov, C., 2004. Does roach behaviour differ between shallow lakes of different environmental state? J. Fish Biol. 65, 135–147. Jakobsen, T.S., Hansen, P.B., Jeppesen, E., Søndergaard, M., 2004. Cascading effect of three-spined stickleback Gasterosteus aculeatus on community composition, size, biomass and diversity of phytoplankton in shallow, eutrophic brackish lagoons. Mar. Ecol. Prog. Ser. 279, 305–309. Juza, T., Kubeˇcka, J., 2007. The efficiency of three fry trawls for sampling the freshwater pelagic fry community. Fish. Res. 85, 285–290. Jeppesen, E., Jensen, J.P., Søndergaard, M., Lauridsen, T.L., Pedersen, L.J., Jensen, L., 1997. Top-down control in freshwater lakes: the role of nutrient state, submerged macrophytes and water depth. Hydrobiologia 342 (343), 151–164. Jeppesen, E., Lauridsen, T.L., Mitchell, S.F., Christoffersen, K., Burns, C.W., 2000. Trophic structure in the pelagial of 25 shallow New Zealand lakes: changes along nutrient and fish gradients. J. Plankt. Res. 22, 951–968. Jeppesen, E., Pekcan-Hekim, Z., Lauridsen, T.L., Søndergaard, M., Jensen, J.P., 2006. Habitat distribution of fish in late summer: changes along a nutrient gradient in Danish lakes. Ecol. Freshw. Fish 15, 180–190. Jørgensen, T.B., 2000. Mossø 1998–1999, tilstand og udvikling. County of Aarhus, in Danish. Koroleff, F. 1970. Determination of total phosphorus in natural water by means of the persulphate oxidation. Cons. Int. Explor. Mer, Interlaboratory Report 3. Lauridsen, T.L., Jeppesen, E., Mitchell, S.F., Lodge, D.M., Burks, R.L., 1999. Diel variation in horizontal distribution of Daphnia and Ceriodaphnia in oligotrophic and mesotrophic lakes with contrasting fish densities. Hydrobiologia 408 (409), 241–250. Lauridsen, T.L., Søndergaard, M., Jensen, J.P., Jeppesen, E., 2005. Undersøgelser i søer – NOVANA. Danmarks Miljøundersøgelser. 234s. Teknisk anvisning fra DMU nr. 22. In Danish. Lewin, W.-C., Okun, N., Mehner, T., 2004. Determinants of the distribution of juvenile fish in the littoral area of a shallow lake. Freshw. Biol. 49, 410–424. ¨ Mehner, T., Holker, F., Kasprzak, P., 2005a. Spatial and temporal heterogeneity of trophic variables in a deep lake as reflected by repeated singular samplings. OIKOS 108, 401–409. ¨ Mehner, T., Diekmann, M., Bramick, U., Lemcke, R., 2005b. Composition of fish communities in German lakes as related to lake morphology, trophic state, shore structure and human use intensity. Freshw. Biol. 50, 70–85. Mehner, T., Holmgreen, K., Lauridsen, T.L., Jeppesen, E., Diekmann, M., 2007. Lake depth and geographical position modify lake fish assemblages of the European ‘Central Plains’ ecoregion. Freshw. Biol. 52, 2285–2297. ¨ Mortensen, E., Jerl Jensen, H., Muller, J.-P., Timmermann, M., 1990. Fiskeundersøgelser i søer. Undersøgelsesprogram, fiskeredskaber og metoder. Danmarks Miljøundersøgelser, 1990. 57 sider. Teknisk anvisning fra DMU, nr. 3. In Danish. Murphy, J., Riley, J.P., 1962. A modified single solution method for determination of phosphate in natural waters. Anal. Chim. Acta 27, 21–26.
T.L. Lauridsen et al. / Fisheries Research 93 (2008) 204–211 ˇ ´ M., Kubeˇcka, J., Hladik, M., Hohausova, ´ E., Cech, ´ J., 2006. Prchalova, M., Frouzova, Fish habitat preferences in an artificial reservoir system. Verh. Internat. Verein. Limnol. 29, 1890–1894. Romare, P., Berg, S., Lauridsen, T., Jeppesen, E., 2003. Spatial and temporal distribution of fish and zooplankton in a shallow lake. Fresw. Biol. 48, 1353–1362. Rowe, D.K., Nichols, S., Kelly, G.R., 2001. Depth distribution and abundance of the common bully, Gobiomorphus cotidianus (Eleotriadae), in three oligotrophic New Zealand lakes, one of which is turbid. Env. Biol. Fish. 61, 407–418.
211
Rudstam, L.G., Lathrop, R.C., Carpenter, S.R., 1993. The rise and fall of a dominant planktivore: direct and indirect effects on zooplankton. Ecology 74, 303–319. Søndergaard, M., Jeppesen, E., Jensen, J.P., Amsinck, S.L., 2005. Water Framework Directive: ecological classification of Danish lakes. J. Appl. Ecol. 42, 616–629. ´ M., Vasek, M., Kubecka, J., Peterka, J., Cech, M., Drastik, V., Hladik, M., Prchalova, Frouzova, J., 2004. Longitudinal and vertical spatial gradients in the distribution of fish within a canyon-shaped reservoir. Internat. Rev. Hydrobiol. 89, 352– 362.
Contents lists available at ScienceDirect
Fisheries Research journal homepage: www.elsevier.com/locate/fishres
A comparison of methods for calculating Catch Per Unit Effort (CPUE) of gill net catches in lakes Torben L. Lauridsen a,∗ , Frank Landkildehus a , Erik Jeppesen a,b , Torben B. Jørgensen a , Martin Søndergaard a a National Environmental Research Institute, University of Aarhus, Department of Freshwater Ecology, Vejlsøvej 25, P.O. Box 314, 8600 Silkeborg, Denmark b University of Aarhus, Department of Biological Sciences, Plant Biology, Ole Worms All´e, Building 135, DK-8000 Aarhus C, Denmark
a r t i c l e
i n f o
Article history: Received 7 February 2008 Received in revised form 22 April 2008 Accepted 22 April 2008 Keywords: Fish Gill nets CEN standard Pelagic Benthic Water Framework Directive
a b s t r a c t Fish constitute an important component of lake ecosystems and many different methods have been used for fish assessment. Based on gill net catches in two stratified (max depth = 14–22 m) eutrophic Danish lakes, relative fish abundance measured as Catch Per Unit Effort (CPUE) was calculated. We used three different methods of which two followed the European standard based on benthic nets (CEN (European Committee for Standardization), 2005: EN 14757. Water quality – Sampling of fish with multi-mesh gill nets. Brussels, 27 pp.), one assuming equal volumes in all depth strata and the other using calculated volumes in the depth strata. The third method followed a modified CEN standard, adopted as a new Danish (DK) standardized method based on calculated benthic and pelagic water volumes and by including both benthic nets and, compared to the CEN standard, an increased fishing effort with pelagic nets. Fish were concentrated in the littoral/benthic part of the upper two depth strata (0–6 m depth) with an up to 8 fold higher abundance than in the pelagic. Calculated CPUE is highly sensitive to the morphometry of the lakes. In lakes with extreme morphometry (unequal volumes in depth strata) it is important to use calculated water volumes for the depth strata. By including information derived from the pelagic nets, total lake CPUEs were 42–56% lower than CPUE values based on benthic nets only. We further show that the relative contribution of CPUE between habitats changes markedly with the nutrient level in 12 deep lakes. It is concluded that for deep lakes it is of key importance to include pelagic nets when comparing fish assemblages and abundances among lakes and when evaluating effects of major changes in key environmental factors, such as nutrient loading and climate. © 2008 Elsevier B.V. All rights reserved.
1. Introduction It is generally accepted that fish strongly influence trophic dynamics in lakes, occasionally with great impact on phytoplankton and on the water quality and ecological state in both freshwater lakes (Carpenter and Kitchell, 1993; Rudstam et al., 1993; Jeppesen et al., 1997, 2000) and brackish lakes (Jakobsen et al., 2004). Fish also contribute to biodiversity and are of interest to both anglers and commercial fishermen. Moreover, fish are one of the four biological variables to be monitored according to the EU Water Framework Directive (WFD) for European water bodies (European Union, 2000). Knowledge of fish community structure and abundance is therefore of key importance. Various gears have been used
∗ Corresponding author. Tel.: +45 89 20 14 00; fax: +45 89 20 14 14. E-mail address: [email protected] (T.L. Lauridsen). 0165-7836/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.fishres.2008.04.007
for fish surveys, such as seine nets, fyke nets, trawls, gill nets, electrofishing and hydroacoustics (e.g., Mehner et al., 2005a; Prchalova´ et al., 2006; Juza and Kubeˇcka, 2007), or assessments have been based on commercial catch data. In general, the methods follow traditions, laws and regulations and the specific sites and aims. However, the use of different gear renders comparisons of fish-data between lakes difficult (Lauridsen et al., 1999; Appelberg, 2000), and even if similar gear has been used, dimensions will most likely differ (Mehner et al., 2007). Results from the marine environment have indicated that an understanding of gear dynamics and environmental influences as well as of habitat and depth influence is important for analyzing CPUE (Bigelow and Maunder, 2007). With the implementation of the EU WFD, it is claimed to use general standards in lake monitoring. Such a standard has not been adopted for fish so far; however, a European standard on gill netting (CEN, 2005) was introduced in 2005. To obtain a standardized Catch Per Unit Effort (CPUE) for fish weight and numbers,
T.L. Lauridsen et al. / Fisheries Research 93 (2008) 204–211
205
Fig. 1. Map of Denmark showing the two study lakes, Lake Brassø and Lake Mossø, in the River Gudena˚ system. Hypsographs illustrate the extreme morphomotry of Lake Brassø and the normal morphometry of Lake Mossø, see text for further explanation.
guidelines for field sampling using a range of depth strata and benthic gill nets combined with pelagic nets at a single site in each lake have been established. In Denmark, a standardized gill netting method (Mortensen et al., 1990) was introduced in 1989 and used in approximately 120 lakes until 2004. The method by Mortensen et al. (1990) prescribed use of benthic, pelagic and floating Lundgreen gill nets (length 42 m, height 1.5 m, multiple mesh with fourteen 3-m panels ranging from 6.25 to 75 mm) together with electrofishing in the littoral zone. The Mortensen et al. (1990) method was not, however, directly comparable to the Swedish and Finnish standards using Nordic gill nets (Appelberg, 2000), which later became the CEN standard (CEN, 2005). To ensure comparability of data, it was decided to substitute the Mortensen et al. (1990) method with the CEN standard (CEN, 2005) for surveys in Danish lakes. However, to increase the effort on pelagic fishes, randomly distributed pelagic gill nets were introduced into the Danish (DK) version of the CEN standard, as pelagic catches may vary considerably from benthic catches and the relative contribution may change with eutrophication (Jeppesen et al., 2006). Such changes may not be registered by just using benthic nets and the change in relative abundance using the CEN standard
Table 1 Characteristics of the two study lakes
Area (ha) Maximum depth (m) Mean depth (m) Retention time (yr) Total phosphorus (g l−1 )
Lake Brassø
Lake Mossø
122 14 4.6 0.01 101 (2005)
1690 22 9 2.2 93 (2005)
is actually not comparable. The DK method allows calculations of CPUE in different ways, of which one is comparable to the CEN standard and the other takes into account the pelagic gill nets. The aim of this study is to demonstrate how the inclusion of pelagic gill nets and the diversion of depth strata into benthic and pelagic zones affect the calculated total fish CPUE in deep lakes. A second aim is to evaluate the necessity of using similar methods in the CPUE data processing.
2. Study area Lake Brassø (122 ha, max. depth 14 m) and Lake Mossø (1690 ha, max depth 22 m) are both situated in the central part of Jutland (Denmark) (Fig. 1). They are part of the River Gudena˚ system and by Danish standard; they are large, eutrophic and deep lakes with low Secchi depth and sparse submerged vegetation. For more morphometric and abiotic characteristics, see Table 1. Since the lakes are large and situated geographically closely in the same river system, their fish stock compositions are very similar (Table 2). Lake Brassø has a deeper part in its eastern end in which the deepest spot covers only a minor area (Fig. 1); a characteristic defined as ‘extreme morphometry’ according to the CEN standard. Lake Mossø has a large deep part in the eastern end but no small deep spot, and its morphemetry is therefore ‘normal’ according to the CEN standard (Fig. 1). In the period June to August and during stratification oxygen concentrations are below 6 mg O2 l−1 in the bottom water (>18 m) of Lake Mossø (Jørgensen, 2000). Lake Brassø is not permanently stratified during summer; however, temporary oxygen depletion does occur in the bottom water (>6 m, T. Jørgensen, unpubl. results).
206
T.L. Lauridsen et al. / Fisheries Research 93 (2008) 204–211
Table 2 List of fish species caught in Lake Mossø and Lake Brassø
Bream, Abramis brama Eel, Anguilla anguillla Gudgeon, Gobio gobio Whitefish, Coregonus spp. Pike, Esox lucius Pikeperch, Sander lucioperca Perch, Perca fluviatilis Roach, Rutilus rutilus Ruffe, Scardinius erythropthalmus Smelt, Osmerus eperlanus Three-spined stickleback, Gasterosteus aculeatus Nine-spined stickleback, Pungitius pungitius
Lake Brassø
Lake Mossø
X X
X X X X X X X X X X X
X X X X X X X X X
3. Material and methods Fishing was performed with Nordic Norm gill nets between August 15 and September 15 according to the DK standard (Lauridsen et al., 2005). According to the CEN standard additional meshes can be added if large fish of certain species are hard to catch (CEN, 2005). Thus, due to the presence of large bream (Abramis brama) in Danish lakes, two extra mesh sizes (68 and 85 mm) were added at the end of the 43 mm mesh of the CEN standard nets, yielding a total of 14 different mesh sizes ranging from 5 to 85 mm knot to knot following a 1.25 geometric series and a total length of 35 m. The sampling procedure is based on stratified random sampling (Appelberg, 2000). The sampled lakes are divided into depth
strata and random sampling is performed within each stratum. By randomizing each gill net within each depth stratum and by randomizing the angle an independent sample was obtained (CEN, 2005). Due to the general shallowness of Danish lakes the DK standard operates with 4 depth strata: 0–3, 3–6, 6–12 and >12 m (Table 3). As both lakes are deeper than 10 m, pelagic nets were added as recommended in the CEN standard. However, following the DK standard, the additional nets are used to undertake an additional stratified random sampling in the pelagic part of each depth stratum, resulting in a minimum of 3 nets in both the pelagic and benthic parts of each stratum (Fig. 2). In the two studied lakes 3–4 benthic and pelagic nets, respectively, were used per stratum. The nets are 1.5 m deep. A standardized fishing period lasted 14–16 h and the nets were set between 4 and 6 p.m. and lifted the following morning between 6 and 8 a.m. Catch Per Unit Effort can be calculated for the total fish stock and for each species. CPUE is defined as a water volume weighted average catch in one net, calculated in three different ways: (1) as whole lake CPUE based on the CEN standard (CEN, 2005) using benthic gill nets and either assuming equal volume in the different depth strata (CEN eqlvol) or (2) using actual volumes in the depth strata (CEN calcvol), or (3) as whole lake CPUE based on the DK standard (Lauridsen et al., 2005) using both benthic and pelagic gill nets and actual volumes in the different parts of the depth strata: CPUEDK =
∗ ∗ ∗ ∗ CPUEben1 + Vpel1 CPUEpel1 + . . . .. + Vben4 CPUEben4 + Vpel4 CPUEpel4 ) (Vben1
Vtotal
Vben1 is the volume of the near-shore area in depth stratum 1; the area between the 0- and the 3-m depth curve multiplied with 1.5
Table 3 Sampling effort in lakes according to the DK (Danish) method (Lauridsen et al., 2005) Maximum depth (m) Area (ha)
Depth strata (m)
−4.5 (# ben)
>4.5–7.5 (# ben/# pel)
>7.5–13.5 (# ben/# pel)
>13.5 (# ben/# pel)
3/3 3
3/3 3/3 x/
3/3 3/3 x/x x 12
< 20
−3* >3–6 >6–12 >12 Tot net
6
6
9
12
20–50
−3* >3–6 >6–12 >12 Tot net
6
3/3 3
3/3 3/3 3
6
9
15
51–100
−3* >3–6 >6–12 >12 Tot net
8
4/4 4
3/3 3/3 3/x
8
12
15
101–250
−3* >3–6 >6–12 >12 Tot net
10
4/4 4
3/3 3/3 3/x
10
12
15
251–1000
−3* >3–6 >6–12 >12 Tot net
12
4/4 4
4/3 4/3 4/x
12
12
18
> 1000
−3* >3–6 >6–12 >12 Tot net
14
6/6 6
4/3 4/3 4/x
18
18
14
3/3 3/3 3/x x/ 15 3/3 3/3 3/x x/ 15 3/3 3/3 3/3 3/x 21 4/3 3/3 3/3 3/x 22 4/3 3/3 3/3 3/x 22
The table shows numbers (#) of nets and how benthic and pelagic nets shall be distributed in the different depth strata at given areas (first column) and given maximum depths (second row). ben, benthic nets; pel, pelagic nets; x, indicates where to place additional nets if maximum lake depth >10 m; (*) if maximum lake depth is ≤4.5 m, depth stratum 1 is expanded to the maximum depth.
T.L. Lauridsen et al. / Fisheries Research 93 (2008) 204–211
Fig. 2. (A) An example of the distribution of benthic and pelagic gill nets in a lake with several depth strata. According to the DK standard (Danish standard) at least 3 nets are set in both the benthic and pelagic part of each depth stratum and (B) division of depth strata volumes into benthic and pelagic parts.
(0.5 × the difference between lower and upper threshold for the depth stratum), and Vpel1 is the corresponding volume in the pelagic part of depth stratum 1; (the area with a depth >3 m) × (the difference between lower and upper threshold for the depth stratum) (Fig. 2). Vtotal is the total lake volume. CPUE can be calculated with and without the 68 mm and 85 mm meshes. In the present study CPUE is calculated without the extra meshes as NPUE (numbers) and WPUE (weight), this being the most frequent standard. The calculations of the depth strata volumes are based on the hypsographs in combination with the bathymetric maps of the lakes (Fig. 1). The latter is used to calculate the areas within the different depth strata. In literature, the percentage of piscivores is suggested as a useful indicator of water quality (Søndergaard et al., 2005). Consequently, the potential piscivore biomass, including all size classes of perch, pike and pikeperch (Stizostedion lucioperca L), measured as percentage (%) of the total fish biomass, is calculated using the three different methods mentioned above. To elucidate how the relative contribution of benthic and pelagic catches changes with the nutrient level, we also compared data collected from 12 deeper lakes (mean depth larger than 6 m), in which only the previously used Mortensen et al. (1990) measure and Lundgreen gill nets were used. Average mean depth of these 12 lakes was 10.2 m and the lake areas ranged between 27 and 1730 ha (Table 4). Since the Mortensen et al. (1990) method did not divide lakes into depth strata (for details see Jeppesen et al., 2006) we calculated average WPUEs for littoral benthic nets and float-
Table 4 Total summer mean phosphorus concentration (TP) and morphometric data on the 12 lakes in which gill netting according to Mortensen et al. (1990) was performed Lake no.
TP (g l−1 )
Mean depth (m)
Max. depth (m)
Area (ha)
1 2 3 4 5 6 7 8 9 10 11 12
114 134 23 135 210 41 210 56 109 32 209 146
10.3 14.6 15 7.1 8 8.7 13.5 13.5 6.3 6.7 9.1 9.9
22 34 33 16 19 15
1301 332 180 315 370 27 1730 941 120 76 1250 662
38 15 14
207
ing nets, respectively, situated 1/2 of the distance from the shore to the lake middle. These results are defined as littoral and pelagic surface results (depth <3 m). Representing the benthic and pelagic of the deeper parts of the lakes we calculated average WPUEs for the benthic nets situated 3/4 of the distance from the shore to the lake middle and for the pelagic nets situated in the middle of the water column and 1/2 of the distance from the shore and the lake middle. These results are defined as benthic and pelagic meta-/hypolimnic samples (depth >6 m). From a pelagial station a depth integrated water sample was collected fortnightly during the period May to October from the epilimnion for total phosphorus (TP). TP was determined as molybdate reactive phosphorus (Murphy and Riley, 1962) following persulphate digestion (Koroleff, 1970). In each lake, habitat effects on total CPUE were tested using a one-way ANOVA on log-transformed fish-data. When the habitat effect was significant, Tukey’s test was applied to data to test for differences between the habitats, between the depth strata and within the depth strata (benthic vs. pelagic nets), at a 0.05 significance level. Similar statistics were applied on the data for the four dominant fish species in the lakes. In the 12 additional lakes, sampled according to Mortensen et al. (1990), we tested for general differences between the littoral and pelagic surface habitats, using a paired t-test on average WPUE values for the two habitats in the 12 lakes. In addition, linear Pearson correlations were performed on benthic littoral, benthic meta-/hypolimnic and pelagic epilimnic WPUE data, respectively, vs. total phosphorus. For unimodal relations, a second order polynomial regression was performed.
4. Results In Lake Brassø there was a significant habitat effect on CPUE (ANOVA test on log-transformed data, p = 0.049 and p = 0.0094 for WPUE and NPUE, respectively), CPUE being highest in the upper depth strata compared to the lowest depth strata. When comparing the benthic catches, WPUE was significantly higher in depth stratum 1 than in 4, and NPUE was significantly higher both in depth strata 1 and 2 than in depth stratum 4 (Tukey’s test, ˛ = 0.05) (Fig. 3A). Within the depth strata, CPUE was higher in the littoral benthic part than in the pelagic part of depth strata 1 and 2 (Fig. 3A); however, the effect was not significant on basis of neither weight nor number (Tukey’s test). In depth strata 3 and 4, CPUE were low and similar between the benthic and pelagic parts (Fig. 3A). In Lake Brassø, 56% of the water volume is in the 1st depth stratum (0–3 m), 29% in the 2nd depth stratum 2 (3–6 m), 14% in the 3rd depth stratum (6–12 m) and 0.5% in the 4th depth stratum (>12 m). Using the CEN eqlvol method, total WPUE and NPUE were 4.51 ± 1.24 (avg ± std) kg net−1 and 146 ± 47 indiv net−1 , respectively (Fig. 3B). For CEN calcvol, WPUE and NPUE were 6.00 ± 1.67 kg net−1 and 200 ± 68 indiv net−1 , respectively (Fig. 3B), and for DK (Fig. 3B) WPUE and NPUE were 2.54 ± 0.63 kg net−1 and 84 ± 22 indiv net−1 (Fig. 3B), respectively. The percentage of the potential piscivore biomass varied between 27 and 32% of the total WPUE (Fig. 3B). Lake Mossø is a lake with “normal” morphometry. Thirty percent of the total water volume is in the 1st depth stratum, and 22, 28 and 20% in the 2nd, 3rd and 4th depth strata, respectively. Like in Lake Brassø, there was a significant habitat effect on CPUE (ANOVA test on log-transformed data, p = 0.035 and p = 0.0006 for WPUE and NPUE, respectively), with CPUE being higher in the littoral benthic part than in the pelagic part in depth strata 1 and 2. However, only in depth stratum 2 did NPUE show a significant difference, and the remaining comparisons were insignificant, both
208
T.L. Lauridsen et al. / Fisheries Research 93 (2008) 204–211
Fig. 3. (A) Weight in g per net (WPUE ± std) and numbers (NPUE ± std) per unit effort in the benthic and pelagic parts of the 4 depth strata in Lake Brassø and (B) average WPUE ± std and NPUE ± std, based on the CEN standard (CEN, 2005), assuming similar water volumes in the depth strata (CEN eqlvol), the CEN standard using calculated water volumes in the depth strata (CEN calcvol) and the DK standard using calculated water volumes and benthic and pelagic gill nets (DK). Percentage of potential piscivore WPUE (perch, pike and pikeperch) is given in numbers.
on weight and number basis (Tukey’s test, ˛ = 0.05) (Fig. 4A). Comparisons of benthic catches showed no significant differences in WPUE between depth strata 1 and 2 and depth stratum 4, but with regard to NPUE the benthic catches were significantly higher in depth strata 1 and 2 compared to depth stratum 4 (Tukey’s test, ˛ = 0.05) (Fig. 4A). Using the CEN eqlvol method, total WPUE and NPUE were 4.27 ± 1.03 kg net−1 and 121 ± 31 indiv net−1 , respectively (Fig. 4B). CEN calcvol gave WPUE and NPUE values of 4.43 ± 1.21 kg net−1 and 124 ± 35 indiv net−1 , respectively (Fig. 4B), and the DK standard, including benthic and pelagic gill nets and calculated water volumes (DK), yielded WPUE and NPUE values of 1.91 ± 0.77 kg net−1 and 45 ± 24 indiv net−1 , respectively (Fig. 4B). Percentage of the potential piscivore biomass varied between 49 and 55% of the total WPUE (Fig. 4B). In both lakes the dominant fish species were perch (Perca fluviatilis), roach (Rutilus rutilus), ruffe (Scardinius erythropthalmus) and bream. Based on the DK method these four species constituted 89% by weight of the total fish stock in Lake Brassø and 64% in Lake Mossø (Fig. 5). In Lake Mossø, perch and roach were evenly distributed, whereas roach was dominant in Lake Brassø, constituting 40% of the total fish stock on weight basis (Fig. 5). The distributional pattern of the four dominant species was similar to the overall fish distribution (Figs. 3 and 4); thus, there were more fish in the benthic parts than in the pelagic parts of the different depth strata (Fig. 5). Except for NPUE of bream in Lake Brassø and WPUE and NPUE of bream in Lake Mossø, there was a significant habitat effect on CPUE (ANOVA test on log-transformed data). In Lake Brassø, comparisons of benthic catches of perch showed significant differences in WPUE between depth strata 1 and 2 and depth stratum 4. With regard to NPUE, a similar result was found. Additionally, the benthic littoral perch-NPUE was significantly higher than the epilimnic pelagic NPUE (Tukey’s test, ˛ = 0.05)
(Fig. 5A). In Lake Brassø, there was no benthic catches of bream in depth strata 3 and 4 (Fig. 5A). Regarding roach, benthic catches in depth strata 1 and 2 were significantly higher compared to NPUE in depth stratum 4 (Tukey’s test, ˛ = 0.05; Fig. 5A). All epilimnic ruffe of Lake Brassø were caught in the benthic littoral part. In depth strata 2 and 3 ruffe were evenly distributed between the benthic and pelagic parts (Fig. 5A). In Lake Mossø, comparisons of benthic catches of perch showed a significantly higher NPUE in depth strata 1 and 2 than in depth strata 3 and 4 (Tukey’s test, ˛ = 0.05) (Fig. 5B). Comparisons of benthic and pelagic catches showed significantly higher benthic NPUE values in depth strata 1 and 2 (Fig. 5B). Bream were caught in very limited numbers and only in the benthic parts of the four depth strata (Fig. 5B). Roach were not caught in depth stratum 4. Comparisons of benthic roach catches showed significantly higher WPUE and NPUE in depth strata 1 and 2 than in depth strata 3 and 4. Due to high variability between replicates no significant differences were observed between benthic and pelagic catches (Fig. 5B). Although the number of ruffe caught was high, the biomass was relatively low. Regarding both WPUE and NPUE, significantly higher abundances were caught in the benthic parts than in the pelagic parts (Tukey’s test, ˛ = 0.05, Fig. 5B). In the 12 lakes sampled according to the previous Danish protocol (Mortensen et al., 1990), littoral WPUE values were in general higher than the pelagic surface WPUE values (p < 0.0001, paired t-test). Additionally, there was a positive correlation between littoral WPUE and total phosphorus, ranging from 3 to 8 kg net−1 in the low-nutrient lakes to 6–13 kg net−1 in the high-nutrient lakes (Fig. 6A). In the meta-/hypolimnic benthic part there was a negative correlation between WPUE and total phosphorus, ranging from 2 to 7 kg net−1 on average in the low-nutrient lakes to 1–5 kg net−1 in the high-nutrient lakes (Fig. 6A). In the corresponding meta/hypolimnic pelagic samples a unimodal relationship was observed, with the highest catch at nutrient levels around 0.1–0.2 mg TP l−1 (Fig. 6B), while lower WPUE values were observed in the pelagic surface water, irrespective the nutrient level (Fig. 6B).
Fig. 4. (A) Weight in g per net (WPUE ± std) and numbers (NPUE ± std) per unit effort in the benthic and pelagic parts of depth strata in Lake Mossø and (B) average WPUE ± std and NPUE ± std. For further explanations, see Fig. 3. Percentage of potential piscivore WPUE is given in numbers.
T.L. Lauridsen et al. / Fisheries Research 93 (2008) 204–211
209
Fig. 5. The percentual biomass contribution of dominant fish species (perch, roach, bream and ruffe) and others (see Table 3 for species) to the total fish stock, and the distribution of the dominant fish species between pelagic and benthic parts of the four depth strata (mean ± std) in Lake Brassø (A) and Lake Mossø (B). The percentual contribution is calculated according to the DK method (for explanation see Fig. 3). ‘*’ Significant at 0.05 level (Tukey’s test).
5. Discussion In the present study more fish were caught in the littoral, benthic parts than in the pelagic zone in the upper two depth strata of Lake Brassø and Lake Mossø. A similar pattern is seen in several previous studies from deep lakes and reservoirs (e.g., Prchalova´ et al., 2006; Fischer and Eckmann, 1997) and shallow lakes (Lewin et al., 2004; Romare et al., 2003). This distribution pattern was also in accordance with the general pattern observed in the 12 lakes sampled with the previous Danish method (Mortensen et al., 1990). A clear difference in fish abundance between the littoral and the pelagic was found particularly in the more nutrient rich lakes. In the two lower depth strata of Lake Brassø and Lake Mossø fish abundance was generally low irrespective of habitat. If oxygen levels and food availability were sufficient we might have expected to encounter more fish at these depths according to the findings in more oligotrophic systems (Prchalova´ et al., 2006; Rowe et al., 2001; Halvorsen et al., 1997). However, the present pattern can be due to limited oxygen levels (Vasek et al., 2004; Jørgensen, 2000) and less food availability (Gliwicz et al., 2006), affecting the presence of fish (Gliwicz et al., 2006; Vasek et al., 2004). This is supported by data from the 12 additional Danish lakes presented in this study, as the highest benthic catches in the lower depth strata are found at low-nutrient levels and reduced catches occur with increasing total phosphorus and probably reduced oxygen levels. Jeppesen et al. (2006) revealed distributional patterns similar to those in the present study as being typical for deeper eutrophic lakes dominated by cyprinid and percid fish species. Except for bream, the four dominant fish species in the two lakes exhibited the same pattern, with high abundance in the benthic parts of the upper depth strata. Particularly in Lake Mossø, the few bream caught were more evenly distributed between the four depth strata (Fig. 5B). The uneven distribution pattern has marked effects on total average CPUE obtained by the different methods used. If equal water volumes were assumed for all depth strata (CEN eqlvol) and
only benthic nets were used, we overestimated total CPUE, as the high CPUE values from the littoral represent the entire depth strata despite the lower actual fish CPUE in the pelagic parts. Also, use of calculated water volumes for the depth strata in combination with benthic nets, which is in accordance with the CEN standard, leads to overestimation of total fish CPUE. According to the CEN standard (CEN, 2005), Lake Brassø has an ‘extreme morphometry’ with large volume variation between the different depth strata. This was clearly evidenced by our results, since CEN calcvol CPUE (calculated volumes) was 33% higher than the CEN eqlvol CPUE (equal volumes). This is basically due to the large proportion of the total lake volume occurring in depth stratum 1 combined with the large CPUE in the littoral. In contrast, Lake Mossø belongs to the category ‘normal morphometry’. The actual volumes in the depth strata are quite similar and, consequently, there was only minor difference between the CEN calcvol CPUE and the CEN eqlvol CPUE (the latter being 3.7% higher). The results underline the importance of using actual volumes when calculating CPUE in lakes, not least when they belong to the ‘extreme morphometry’ category. In the CEN standard, it is assumed that benthic net catches represent all depth strata, but this seems not to be true for Lake Brassø and Lake Mossø, as overall lower CPUE was obtained in the pelagic than in the benthic parts. Consequently, total CPUE calculated according to the DK method for Lake Brassø was 42–56% lower than when using the official CEN CPUE method, and 55% lower for Lake Mossø. We found that the CPUE based percentage piscivore biomass is almost similar in the two study lakes irrespective of the method used, which is important if this measure is to be used as a water quality indicator (Søndergaard et al., 2005). It can be claimed that the bias introduced by using the official CEN method is similar and that the data from different lakes are thus comparable. However, the data from the additional 12 lakes and an earlier study by Jeppesen et al. (2006) show that the differences between the two methods depend on the nutrient level
210
T.L. Lauridsen et al. / Fisheries Research 93 (2008) 204–211
Acknowledgements We would like to thank the Danish counties and in particular County of Aarhus for access to data, Anne Mette Poulsen for manuscript assistance and Tinna Christensen and Juana Jacobsen for layout assistance. This work was financially supported by the EU EuroLimpacs project (GOCE-CT-2003-505540) and the CLEAR project (a Villum Kann Rasmussen Centre of Excellence Project). References
Fig. 6. Data represent 12 deeper lakes (mean depth >6 m) and encompass 1–3 lake years. Weight per unit effort (WPUE ± std) in (A) benthic nets from the littoral (LitBen, depth <3 m, n = 12–24) and 3/4 of the distance from the shore to the lake middle (HypoBen, depth >6 m, n = 4–12) and (B) floating nets (Epi, n = 5–6) situated 1/2 of the distance from the shore to the lake middle and pelagic meta-/hypolimnic nets (Hypo) situated in the middle of the water column and 1/2 of the distance from the shore to the lake middle (depth >6 m, n = 6–12). Pearson correlation coefficients and a second order polynomial regression coefficient (on the unimodal response) are given.
and are lower in the more oligotrophic systems. Consequently, an evaluation of the effects of changes in nutrient loading cannot be done properly without considering also the pelagic net data. Moreover, species richness and diversity are higher in benthic and littoral habitats compared to pelagic habitats (Diekmann et al., 2005; Mehner et al., 2005b), which suggests that the CEN standard will not give a true picture of the proportion of the different fish species in the lake. Furthermore, the spatial distribution and survival of fish are strongly influenced by the temperature and oxygen content. Vendace (Coregonus albula), for instance, prefer cold hypolimnic water and may also display habitat shifts if conditions change, e.g., the oxygen content (George et al., 2006). Such a habitat shift may wrongly be interpreted as a reduction in vendace CPUE according to the CEN standard. Conditions may also change with alterations in influencing stressors such as climate and nutrients. In Danish lakes, for example, increasing abundances of, for instance, roach and perch occur in the littoral with increasing total phosphorus concentrations (Jeppesen et al., 2006). Consequently, a reduction in nutrient loading may lead to a behavioural change towards the pelagic (Jacobsen et al., 2004; Jeppesen et al., 2006), which could be detected in the CEN standard as reduced CPUE. We therefore emphasize the importance of including pelagic data. Our results are based on relatively few lakes, but underline the importance of using a common method when comparing CPUE values between lakes and regions.
Appelberg, M. (Ed.), 2000. Swedish standard methods for sampling freshwater fish with multi-mesh gillnets. Fiskeriverket information, 1. Bigelow, K.A., Maunder, M.N., 2007. Does habitat or depth influence catch rates of pelagic species? Can. J. Fish. Aquat. Sci. 64, 1581–1594. Carpenter, S.R., Kitchell, J.F. (Eds.), 1993. The Trophic Cascade in Lakes. Cambridge University Press, New York. CEN (European Committee for Standardization), 2005: EN 14757. Water quality – Sampling of fish with multi-mesh gill nets. Brussels, 27 pp. ¨ Diekmann, M., Bramick, U., Lemcke, R., Mehner, T., 2005. Habitat-specific fishing revealed distinct indicator species in German lowland lake fish communities. J. Appl. Ecol. 42, 901–909. European Union, 2000. Directive 2000/60/EC of the European Paliament and of the Council Establishment a Framework of the Community Action in the Field of Water Policy. European Commision, off. J. Eur. Commun. L327, 1. Fischer, P., Eckmann, R., 1997. Spatial distribution of littoral fish species in a large European lake, Lake Constance, Germany. Arch. Hydrobiol. 140, 91–116. George, D.G., Bell, V.A., Parker, J., Moore, R.J., 2006. Using a 1-D mixing model to assess the potential impact of year-to-year changes in weather on the habitat of vendace (Coregonus albula) in Bassenthwaite Lake, Cumbria. Freshw. Biol. 51, 1407–1416. Gliwicz, Z.M., Slon, J., Szynkarczyk, I., 2006. Trading safety for food: evidence from gut contents in roach and bleak captured at different distances offshore from their daytime littoral refuge. Freshw. Biol. 51, 823–839. Halvorsen, M., Jørgensen, L., Amundsen, P.-A., 1997. Habitat utilization of juvenile Atlantic trout (Salmo trutta L.) and Arctic charr (Salvelinus alpinus L.) in two lakes in northern Norway. Ecol. Freshw. Fish 6, 67–77. Jacobsen, L., Berg, S., Jepsen, N., Skov, C., 2004. Does roach behaviour differ between shallow lakes of different environmental state? J. Fish Biol. 65, 135–147. Jakobsen, T.S., Hansen, P.B., Jeppesen, E., Søndergaard, M., 2004. Cascading effect of three-spined stickleback Gasterosteus aculeatus on community composition, size, biomass and diversity of phytoplankton in shallow, eutrophic brackish lagoons. Mar. Ecol. Prog. Ser. 279, 305–309. Juza, T., Kubeˇcka, J., 2007. The efficiency of three fry trawls for sampling the freshwater pelagic fry community. Fish. Res. 85, 285–290. Jeppesen, E., Jensen, J.P., Søndergaard, M., Lauridsen, T.L., Pedersen, L.J., Jensen, L., 1997. Top-down control in freshwater lakes: the role of nutrient state, submerged macrophytes and water depth. Hydrobiologia 342 (343), 151–164. Jeppesen, E., Lauridsen, T.L., Mitchell, S.F., Christoffersen, K., Burns, C.W., 2000. Trophic structure in the pelagial of 25 shallow New Zealand lakes: changes along nutrient and fish gradients. J. Plankt. Res. 22, 951–968. Jeppesen, E., Pekcan-Hekim, Z., Lauridsen, T.L., Søndergaard, M., Jensen, J.P., 2006. Habitat distribution of fish in late summer: changes along a nutrient gradient in Danish lakes. Ecol. Freshw. Fish 15, 180–190. Jørgensen, T.B., 2000. Mossø 1998–1999, tilstand og udvikling. County of Aarhus, in Danish. Koroleff, F. 1970. Determination of total phosphorus in natural water by means of the persulphate oxidation. Cons. Int. Explor. Mer, Interlaboratory Report 3. Lauridsen, T.L., Jeppesen, E., Mitchell, S.F., Lodge, D.M., Burks, R.L., 1999. Diel variation in horizontal distribution of Daphnia and Ceriodaphnia in oligotrophic and mesotrophic lakes with contrasting fish densities. Hydrobiologia 408 (409), 241–250. Lauridsen, T.L., Søndergaard, M., Jensen, J.P., Jeppesen, E., 2005. Undersøgelser i søer – NOVANA. Danmarks Miljøundersøgelser. 234s. Teknisk anvisning fra DMU nr. 22. In Danish. Lewin, W.-C., Okun, N., Mehner, T., 2004. Determinants of the distribution of juvenile fish in the littoral area of a shallow lake. Freshw. Biol. 49, 410–424. ¨ Mehner, T., Holker, F., Kasprzak, P., 2005a. Spatial and temporal heterogeneity of trophic variables in a deep lake as reflected by repeated singular samplings. OIKOS 108, 401–409. ¨ Mehner, T., Diekmann, M., Bramick, U., Lemcke, R., 2005b. Composition of fish communities in German lakes as related to lake morphology, trophic state, shore structure and human use intensity. Freshw. Biol. 50, 70–85. Mehner, T., Holmgreen, K., Lauridsen, T.L., Jeppesen, E., Diekmann, M., 2007. Lake depth and geographical position modify lake fish assemblages of the European ‘Central Plains’ ecoregion. Freshw. Biol. 52, 2285–2297. ¨ Mortensen, E., Jerl Jensen, H., Muller, J.-P., Timmermann, M., 1990. Fiskeundersøgelser i søer. Undersøgelsesprogram, fiskeredskaber og metoder. Danmarks Miljøundersøgelser, 1990. 57 sider. Teknisk anvisning fra DMU, nr. 3. In Danish. Murphy, J., Riley, J.P., 1962. A modified single solution method for determination of phosphate in natural waters. Anal. Chim. Acta 27, 21–26.
T.L. Lauridsen et al. / Fisheries Research 93 (2008) 204–211 ˇ ´ M., Kubeˇcka, J., Hladik, M., Hohausova, ´ E., Cech, ´ J., 2006. Prchalova, M., Frouzova, Fish habitat preferences in an artificial reservoir system. Verh. Internat. Verein. Limnol. 29, 1890–1894. Romare, P., Berg, S., Lauridsen, T., Jeppesen, E., 2003. Spatial and temporal distribution of fish and zooplankton in a shallow lake. Fresw. Biol. 48, 1353–1362. Rowe, D.K., Nichols, S., Kelly, G.R., 2001. Depth distribution and abundance of the common bully, Gobiomorphus cotidianus (Eleotriadae), in three oligotrophic New Zealand lakes, one of which is turbid. Env. Biol. Fish. 61, 407–418.
211
Rudstam, L.G., Lathrop, R.C., Carpenter, S.R., 1993. The rise and fall of a dominant planktivore: direct and indirect effects on zooplankton. Ecology 74, 303–319. Søndergaard, M., Jeppesen, E., Jensen, J.P., Amsinck, S.L., 2005. Water Framework Directive: ecological classification of Danish lakes. J. Appl. Ecol. 42, 616–629. ´ M., Vasek, M., Kubecka, J., Peterka, J., Cech, M., Drastik, V., Hladik, M., Prchalova, Frouzova, J., 2004. Longitudinal and vertical spatial gradients in the distribution of fish within a canyon-shaped reservoir. Internat. Rev. Hydrobiol. 89, 352– 362.