The effect of fishing effort, fish stocking, and population density of overwintering cormorants on the harvest and recapture rates of three rheophilic fish species in central Europe

Fisheries Research 223 (2020) 105440

Contents lists available at ScienceDirect

Fisheries Research journal homepage: www.elsevier.com/locate/fishres

The effect of fishing effort, fish stocking, and population density of overwintering cormorants on the harvest and recapture rates of three rheophilic fish species in central Europe

T

Roman Lyach1 Institute for Evaluations and Social Analyses, 186 00 Prague, Czech Republic

ARTICLE INFO

ABSTRACT

Handled by S. Xavier Cadrin

Rheophilic fish species are among the most threatened groups of organisms. There is evidence that anthropogenic activities, such as fish stocking and fishing pressure, could have a significantly negative effect on freshwater fish populations. In addition to this, high population densities of cormorants, Phalacrocorax carbo, (one of the most important avian piscivorous predators in Europe) are also partially driven by anthropogenic activities. The first goal of this study was to estimate the effects of fishery management on fish harvest in three rheophilic fish species: barbel (Barbus barbus), nase (Chondrostoma nasus), and vimba bream (Vimba vimba). The second goal was to estimate the effect of cormorant population density on the harvest of these species. Individual mandatory angling logbooks data were collected on the 22 largest rivers in central Bohemia (Czechia, central Europe) over 13 years. Cormorant density was estimated using bird census data while the diet composition of cormorants was assessed using an analysis of fish diagnostic bones extracted from regurgitated bird pellets. In total, anglers visited the selected fishing areas 3 million times over 13 years and harvested 1 000 tons of fish, of which 4.5 tons were rheophilic species. The fishing effort was the most important driver of fish harvest. Fish stocking and cormorant population density did not significantly affect the fish harvest. However, rheophilic fish species comprised 8.8 % of fish consumed by cormorants (by biomass). In conclusion, fishing pressure was the most important factor affecting the harvest of rheophilic fish. Conversely, the intensity of predation by cormorants was not significantly affecting fish harvest despite their consumption of rheophilic fish. However, there is a possibility that predation pressure of cormorants was evenly distributed among fishing areas.

Keywords: Angling diary Aquaculture conflict Bird predation Fisheries management Sport fishing

1. Introduction Freshwater fish species are among the most threatened groups of animals worldwide (FAO, 2010). Rheophilic fish species are among the most important and abundant vertebrates in medium sized and larger rivers. Over the last few decades, the number of rheophilic fish has decreased throughout Europe (Penczak et al., 1998; Mueller et al., 2018). The main reasons for these population decreases are overexploitation by uncontrolled fish harvesting, water pollution (organic pollutants and pharmaceutical residuals), the construction of river obstacles (weirs, dams, hydropower plants), weather extremes (floods, droughts), and also greater predation pressure caused by increasing numbers of fish-eating birds and mammals (cormorant Phalacrocorax sp., otters Lutra sp., mink Mustela vison, herons Ardea sp.) (Suter, 1995; Dudgeon et al., 2006; Lajus et al., 2013; Tonolla et al., 2017; Piria et al., 2019).

1

Fishing pressure can negatively affect rheophilic fish populations by removing larger fish which are usually highly fertile females with high fitness and potential for reproduction (Birkeland et al., 2015a,b; Gwinn et al., 2015). Even anglers who prefer the increasingly popular catchand-release fishing strategy may be responsible for killing 10 %–20 % of released fish because of post-release mortality (Bartholomew and Bohnsack, 2005). This mortality also applies to smaller fish that have not yet reproduced. Fish stocking is supposed to support and bolster wild fish populations and also provide catches for recreational anglers. However, studies have shown that fish stocking is often ineffective and does not significantly affect harvest per effort (e.g. Yule et al., 2000; Baird et al., 2006; Michaletz et al., 2008). Nevertheless, some studies have shown partially positive effects of fish stocking on fish harvest (Wiley et al., 1993). The effect of cormorant, Phalacrocorax carbo, predation on the

E-mail address: [email protected]. ResearchGate: https://www.researchgate.net/profile/Roman_Lyach2

https://doi.org/10.1016/j.fishres.2019.105440 Received 3 July 2019; Received in revised form 12 November 2019; Accepted 13 November 2019 0165-7836/ © 2019 Elsevier B.V. All rights reserved.

Fisheries Research 223 (2020) 105440

R. Lyach

harvest of rheophilic fish has been studied with warrying results. By studying cormorant diet in central Europe, researchers have discovered that cormorants frequently prey on shoaling fish species which qualifies rheophilic fish as preferable prey (Suter, 1995, 1997; Keller, 1998). Some researchers claim that large numbers of cormorants hunting on fishing grounds lead to decreased harvests in both recreational and commercial fisheries (Jepsen et al., 2018; Takai et al., 2018). Other studies did not reveal a significant link between cormorant numbers and fish harvest (Čech and Vejřík, 2011a; Lehikoinen et al., 2017). There are several studies that have tried to determine the effects of fish stocking, fishing effort, or cormorant predation on the harvest of rheophilic fish species (Keckeis et al., 1997; Florea, 2008; Golski et al., 2017; Trigo et al., 2017; Brinker et al., 2018; Schletterer et al., 2018). However, there is no study that describes the effect of all three factors on the harvest of rheophilic fish species in central Europe on a larger spatio-temporal scale. This study analysed basic factors in fisheries over 13 years on 22 inland fishing areas (river stretches) that connect 11 500 km2 of land and have a water surface area of 15.9 km2. The goal of this study was to fill this knowledge gap by describing the effects of fish stocking, fishing effort, and cormorant population density on the harvest and recapture rates of three rheophilic fish species: barbel, Barbus barbus, nase Chondrostoma nasus, and vimba bream, Vimba vimba. This study had three main aims: Firstly, it aimed to describe changes over time in three metrics: (a) the harvest and stocking practices of rheophilic fish, (b) the representation of rheophilic fish in overall harvest, and (c) the numbers of overwintering cormorants in central Bohemia. Secondly, it aimed to assess the effects of fish stocking and fishing effort on the harvest and recapture of stocked rheophilic fish species. Fish stocking is performed to increase fish harvest and recapture rates. Therefore, it was expected that a higher intensity of fish stocking and fishing effort would lead to a higher harvest and recapture of rheophilic fish. Thirdly and lastly, it aimed to assess the diet composition of resident cormorants and connect cormorant population densities to the harvest and recapture of stocked fish. Cormorants prefer to prey on shoaling rheophilic fish species. Therefore, it was expected that higher cormorant population densities would be connected to the lower harvest and recapture of stocked fish.

2.2. Recreational fishing in Czechia Recreational fishing in Czechia is organized by the Czech Fishing Union (the main authority in recreational fishing in Czechia) and is centralized for the whole country. Anglers are required to record all fishing visits and harvested (killed) fish into individual angling logbooks. Fish that are undersized, caught during a closed season, or otherwise released are not recorded in logbooks. Completed logbooks are collected at the end of each year (or in January of the following year). Anglers are obliged to submit logbooks in order to receive a new empty logbook for the next year. Therefore, return rates of filled in logbooks are above 99 %. Fishing areas are divided into two categories: salmonid and non-salmonid fishing areas. Salmonid fishing areas are defined as smaller streams and rivers, while non-salmonid fishing areas are defined as larger rivers. Salmonid fishing areas should provide good environmental conditions for natural salmonid growth and reproduction. Non-salmonid fishing areas are defined as fishing areas that do not meet the criteria of salmonid fishing areas. Fishing areas are defined as stretches of rivers or streams that are divided by obstacles or buildings (weirs, dams, hydropower plants, bridges). The size (water surface area) of the studied fishing areas was 18–150 ha and the average size of a fishing area was 32 ha. For detailed description of recreational fishing in Czechia, see Lyach and Čech (2018). 2.3. Angling rules for rheophilic fish species All three studied rheophilic fish species (barbel, Barbus barbus, nase, Chondrostoma nasus, and vimba bream, Vimba vimba) are species of medium value in recreational fishing. All rheophilic species have a combined bag limit of 7 kg of fish per angler per day. Whenever an angler reaches or exceeds this limit, the angler is obliged to stop fishing for that day. This rule is applied for all fishing areas under jurisdiction of the Czech Fishing Union (about 80 % of all fishing areas in Czechia and all non-private fishing areas in the study area). The limit of 7 kg can be exceeded (e.g. killing three fish each weighing 3 kg is legal). All caught fish must be noted in individual angling logbooks, including the date of catch, size of fish [cm and kg], and the ID of the fishing reach. The minimum legal angling sizes for rheophilic species are the following: nase (30 cm TL, total length), barbel (40 cm TL, total length), and vimba bream (25 cm TL, total length). All three species have a closed season from 16 March to 15 June. Fish that are either too small or are caught during closed season have to be returned back to the water. Anglers are obliged to measure kept fish to the nearest cm and note the catch into their angling logbooks. Fish are measured in total length (TL in cm). The listed fishing regulations did not change between years 2005–2017 and were effective for all fishing areas in the study area.

2. Material and methods 2.1. Study area This study was carried out in the region of central Bohemia (49.5°–50.5 °N, 13.5°–15.5 °E), Czechia (central Europe, Fig. 1). The region covers an area of 11 015 km2 and has a mostly rural and agricultural character. The study area is dominated by the rivers Elbe and Vltava. Both rivers belong to the upper Elbe River Basin. All rivers in the study area belong to the North Sea Drainage area. The studied fishing areas are situated in lowlands with an altitude of 200–600 m above sea level. Waters in the study areas are mostly mesotrophic and eutrophic with a biomass of 150–300 kg of fish per ha (Lyach and Čech, 2018). The study area includes salmonid streams (smaller streams, mostly less than 10 m wide, usually dominated by salmonids) and nonsalmonid rivers (wider streams and rivers, usually 10–300 m wide, dominated by cyprinids or percids). The majority of rivers and streams in the study area have natural reproducing populations of rheophilic fish species (Czech Fishing Union, unpubl. data). Fish populations in the studied rivers and streams seem to be in relatively good condition due to natural fish reproduction and intensive fish stocking (Czech Fishing Union, unpubl. data). Cormorants roosted and hunted in the area from October to March (Czech Fishing Union, unpubl. data). Two of the studied species (barbel, Barbus barbus, and vimba bream, Vimba vimba,) are native to the study area while nase, Chondrostoma nasus, is a nonnative species.

2.4. Stocking of rheophilic fish species Rheophilic fish species are frequently stocked in the study area every year. Barbel, nase, and vimba bream are mostly stocked in larger rivers (50–300 m wide). Fishery management stocks fish of a wide variety of sizes, including small 0, 0+, and YOY (young-of-the-year) fish (5–10 cm TL) and also a small amount of larger fish (10–30 cm TL). Small fish are usually stocked in hundreds or thousands per one river or stream. Larger fish are usually stocked in tens or hundreds of kilograms per one hectare of a river (i.e. tens or hundreds of fish per hectare). The main purpose of fish stocking is to support naturally reproducing fish populations, however, larger fish are stocked for angling purposes as well. Before fish stocking occurs, all fish are weighed together to the nearest 100 g. Smaller fish are weighed all together in one bag, larger fish are weighed in groups of 40–100. The number of stocked fish is then estimated from the overall weight by applying length-weight equations of the specific fish species. The length-weight equations are based on catch data of a larger amount (at least 1 000) of fish that were 2

Fisheries Research 223 (2020) 105440

R. Lyach

Fig. 1. A map of the study area with the regions of Prague and central Bohemia (Czechia, central Europe) highlighted where fishing areas (marked rivers) and cormorant roosting sites (full black rectangle) were situated.

caught in the study area by the managers of fisheries. Fish stocking is performed by managers of local fisheries and monitored by the Czech Fishing Union. Since the Czech Fishing Union is the only authority that sells/collects/manages fishing permits and benefits from fish stocking on studied rivers and streams, no other subjects perform fishery management (especially fish stocking) on the studied waters.

under a stereo microscope (magnification 8–16 ×). Fish species were identified to the lowest possible taxonomic level based on morphological differences of diagnostic bones (os pharyngeum, maxillare, dentale, intermaxillare, operculare, praeoperculare, praevomer, cleithrum, basioccipitale, mesethmoid, vertebra, coracoid, vomer and pectoral spine). All diagnostic bones found in regurgitated pellets were measured to the nearest 0.1 mm and paired with each individual pellet whenever possible. The number of fish represented in a pellet was determined by the highest total of any identifiable part. Our own collection of diagnostic bones was used to determine the original size of worn-down bones. The length of worn-down bones was estimated by comparing the least worn parts of the extracted bone to the same parts of fully preserved bones from our own collection. This collection was also used in previous studies that analysed the diet of fish-eating predators (Čech et al., 2008; Čech and Vejřík, 2011b; Čech and Čech, 2013, 2017; Lyach et al., 2018). The length of fully preserved diagnostic bones was then used for calculating the original prey fish length (TL, total length in cm), using length-length equations from studies by Čech et al. (2008); Čech and Vejřík (2011a, b), Čech and Čech (2013, 2017), and Lyach et al. (2018). Prey fish weight was calculated using length-weight equations from the work of Čech and Čech (2017) and from FishBase (FishBase.org).

2.5. Cormorant census Cormorants were counted on 22 of the largest and most important fishing areas (river stretches) in the study area. Previously trained bird watchers (n = 34) counted roosting, hunting, sleeping, and feeding cormorants that were present above water, circling in close proximity to the river, or roosting on trees near the water bank. The birds were counted over winter 2017/2018 during the following dates: 21–22 October, 18–19 November, 16–17 December, 20–21 January, 17–18 February, 17–18 March. 2.6. Cormorant diet analysis Cormorant diet was investigated using an analysis of regurgitated pellets (Barrett et al., 2007). Pellets were collected during eight visits over winters 2014/2015 and 2015/2016 (four visits each winter in November, December, January, and February) at roosting places in Central Bohemia during daylight (late morning till midday), when few birds were present in the area. About 500 m2 of the ground was searched for pellets during each visit. All available pellets were collected during each visit and put individually into plastic bags into deep freeze (−18 °C). In a laboratory, each pellet was mixed in a solution of 300 ml warm tap water (50 °C) and 15 g of 1 M, 97–99 % Na(OH) for approximately one minute until the gastric mucosa was completely dissolved. The remaining hard parts were washed through a sieve (mesh size 0.5 mm) to remove the remaining dirt, mucosa, and Na (OH). The cleaned hard parts were then put into Petri dishes. All recognizable hard parts were separated, dried at room temperature, and analysed

2.7. Data sources Data from annual angling summaries of fish catches were used for the purpose of this study. These data originated from mandatory angling logbooks collected from individual anglers. For an example of an angling summary and angling logbook, see Lyach and Čech (2018). Data from 22 inland freshwater fishing areas over the course of years 2005–2017 were used. The selected fishing areas covered a water surface area of 15.9 km2. The data were originally collected by the Czech Fishing Union and later processed by the author of this study. Fishing areas are defined as stream and river stretches where recreational fishing can be legally conducted. Similar datasets have been previously 3

Fisheries Research 223 (2020) 105440

R. Lyach

used for scientific purposes (Humpl et al., 2009; Jankovský et al., 2011; Boukal et al., 2012; Lyach and Čech, 2018).

Anglers visited the locations included in this study on 3 027 612 occasions and harvested 1 022 388 kg of fish, among which barbel comprised 2 651 kg, nase 583 kg, and vimba bream 1 368 kg. Fishery managers stocked 601 kg of barbel, 252 kg of nase, and 181 kg of vimba bream. Barbel formed 0.28 % of the overall harvest while nase formed 0.06 % and vimba bream formed 0.14 %. The recapture rates of stocked fish were 3.8 % for barbel, 0.06 % for nase, and 1.14 % for vimba bream. Roach, Rutilus rutilus, dominated the diet of cormorants (50 % and 60 % by number and biomass, receptively), while common carp, Cyprinus carpio, dominated in the harvest of anglers (75 % and 85 % by number and biomass, respectively).

2.8. Measured metrics This study assessed the following metrics: harvest per fishing visit [fish biomass in kg], harvest per one hectare of fishing areas [fish biomass in kg], intensity of fish stocking per hectare of fishing areas [fish biomass in kg], the recapture (harvest) rates of stocked fish [the biomass of harvested fish divided by the biomass of stocked fish, results are in %], the percentage of a specific rheophilic fish species in the overall fish harvest [%], the number of cormorants per hectare of fishing areas (birds per hectare), and fishing effort (number of fishing visits per year). To estimate the effect of fish stocking intensity on fish harvest, data on fish stocking were used exclusively from 3 to 5 years prior harvest. The mean value of three consecutive years was used in the analysis. Data on fish stocking from 0 to 2 years prior the harvest were excluded because the stocked fish were mostly small (5–10 cm TL, total length) and unlikely to grow to legal angling size (25–40 cm TL, total length) over two years. Data on fish stocking older than 5 years were excluded for two reasons: (1) the usual lifespan of the studied rheophilic fish in the wild is 5–6 years maximum, and (2) the stocked fish usually display high mortality due to stocking stress, predation, angling, and an inability to adapt to natural conditions.

3.2. Changes over time Several basic parameters in recreational fisheries significantly changed over time between the years 2005 and 2017. The amount of migrating and roosting cormorants in the area increased over time (W = 39 168, p = 0.01) as well as the number of anglers (W = 33 336, p = 0.03). Conversely, the amount of fish harvested by anglers decreased over time – this was true for all fish species pooled together (W = 13 860, p < 0.01) as well as for barbel (W = 33 696, p = 0.04), nase (36 782, p = 0.02), and vimba bream (W = 37 152, p = 0.02). On the other hand, fish stocking practices did not change over time – fishery managers stocked similar quantities of rheophilic fish every year. This was true for barbel (W = 24 352, p = 0.50), nase (24 568, p = 0.18), and vimba bream (W = 22 768, 0.20).

2.9. Statistical analysis The statistical programme R (R i386 3.4.1., R Development Core Team 2017) was used for statistical testing. The package for generalized linear models (GLMM) was used to fit the models (Hadfield, 2010). The function lmer in the package lme4 (version 0.999375‐42; Bates et al., 2015) was used to calculate R-squared values (Nakagawa et al., 2017). Akaike information criterion (AIC) was used to compare the fit of the models used. One fishing area was the base sampling unit and was treated as a random effect because no information was available on angler preferences of fishing areas. Fish stocking, fishing effort, and cormorant population density were treated as fixed effects. Harvest per hectare and effort, recapture rates, and representation of rheophilic fish in the overall harvest were treated as response variables. Gamma error distribution with log link function was used in the models. The equation for models was the following: harvest ∼ stocked fish + cormorant population density + fishing effort + (1|fishing reach). Data distributions were tested using Shapiro-Wilk tests while differences between two groups were tested using Wilcoxon tests. A probability level of up to p = 0.05 was accepted for all the statistical tests, and all statistical tests were two-tailed. Bonferroni correction was applied when multiple groups were compared in statistical analysis. The results presented in the tables were derived from models calculated using R statistical programme while the figures were drawn in MS Excel. The above described method was previously used to analyse similar data sets (Humpl et al., 2009; Jankovský et al., 2011; Boukal et al., 2012; Lyach and Čech, 2018).

3.3. Harvest per effort Harvests of rheophilic fish species were mostly affected by fishing effort (Table 1). Fishing areas with a higher fishing effort displayed lower harvest per effort values in all three species: barbel, nase, and vimba bream (Fig. 2). The harvest rates of rheophilic fish species were not significantly affected by the intensity of fish stocking (Table 1). Similarly, population densities of cormorants in the area had no significant effect on the harvest per effort of these rheophilic fish. 3.4. Harvest per hectare Harvest per hectare was significantly affected by fishing effort (Table 1, Fig. 3). However, this was true only for nase and vimba bream; no relationship was found between harvest of barbel and fishing effort. When the fishing areas were divided into two groups based on the intensity of fishing effort, it was discovered that the results in both groups were different. Fishing areas with a low fishing effort mostly showed a positive effect of fishing effort on harvest per hectare. Conversely, fishing areas with a high intensity of fishing effort showed a negative effect of fishing effort on harvest per hectare. In all occasions, neither fish stocking nor population densities of cormorants had any significant effect on harvest per hectare. 3.5. Representation of rheophilic fish in overall harvest

3. Results

The representation (percentage) of rheophilic fish in the overall fish harvest was mostly affected by fishing effort (Table 1). Similarly to harvest per hectare, when the fishing areas were divided into two groups based on intensity of fishing effort, it was discovered that the relationships were different in both groups (Fig. 4). Fishing areas with a low fishing effort mostly showed a positive effect of fishing effort on the representation of rheophilic fish in the overall fish harvest. Conversely, fishing areas with a high intensity of fishing effort showed a negative effect of fishing effort on the representation of rheophilic fish in the overall fish harvest. Similarly to harvest per effort and per hectare, fish stocking and cormorant population density had no significant effect on the representation of rheophilic fish in the overall fish harvest.

3.1. A summary of data over 13 years The average number of observed cormorants on the studied fishing areas was 50 birds per fishing reach (ranging between 0–300 birds) and 0.56 birds per hectare of water surface (ranging between 0–2.88 birds). Zero cormorants were observed at several fishing areas from November to March. On the other hand, approximately 300 cormorants were observed at one section of the Vltava River (the largest river in Czechia) during November. More cormorants per hectare were observed at fishing areas situated on larger rivers (the Vltava and Elbe Rivers) in comparison to smaller rivers. 4

Fisheries Research 223 (2020) 105440

R. Lyach

Table 1 Results of models used in this study, showing the best fit using fixed effects (fishing effort, fish stocking, and cormorant population density) and random effect (fishing reach). The upper and lower table shows models that include data from fishing reaches with lower fishing visit rates (1 000–18 000 visiting anglers per year) and higher fishing visit rates (35 000–55 000 visiting anglers per year), respectively. Note: SD = standard deviation, AIC = Akaike information criterion, DF = degrees of freedom. reponse variable

intercept

a) fishing areas with lower fishing visit rates (1 000–18 harvest per effort: barbel 6.33E-05 harvest per effort: nase 2.24E-06 harvest per effort: vimba bream 2.43E-05 recapture rates: barbel 1.65E+00 recapture rates: nase −2.60E-01 recapture rates: vimba bream 6.24E-01 % of barbel in harvest 6.56E-04 % of nase in harvest −9.34E-03 % of vimba bream in harvest 4.31E-02 harvest per hectare: barbel 7.67E-02 harvest per hectare: nase −9.34E-03 harvest per hectare: vimba bream 4.31E-02

SD (intercept)

fixed variable

000 visiting anglers per year) 8.05E-06 fishing effort 5.76E-07 fishing effort 3.77E-06 fishing effort 2.09E+00 fishing effort 9.72E-02 fishing effort 1.24E+00 fishing effort 2.06E-04 fishing effort 4.52E-03 fishing effort 1.14E-02 fishing effort 1.16E-02 fishing effort 4.52E-03 fishing effort 1.14E-02 fishing effort

b) fishing areas higher fishing visit rates (35 000–55 000 visiting anglers per year) harvest per effort: barbel 4.14E-05 2.73E-05 harvest per effort: nase 3.64E-06 2.15E-06 harvest per effort: vimba bream 4.68E-05 1.94E-05 recapture rates: barbel 2.57E+00 1.80E+01 recapture rates: nase 2.21E-01 1.64E-16 recapture rates: vimba bream 2.86E+00 7.04E-01 % of barbel in harvest 4.46E-02 1.97E-02 % of nase in harvest 4.90E-03 7.68E-03 % of vimba bream in harvest 3.21E-02 1.24E-02 harvest per hectare: barbel 1.61E+00 1.15E+00 harvest per hectare: nase 3.24E-01 5.07E-01 harvest per hectare: vimba bream 1.75E+00 7.58E-01

fishing fishing fishing fishing fishing fishing fishing fishing fishing fishing fishing fishing

effort effort effort effort effort effort effort effort effort effort effort effort

slope

SD (slope)

p-value

R-squared

AIC

DF

−4.34E-09 −1.59E-10 −1.58E-09 3.53E-04 5.62E-05 −4.29E-05 6.66E-08 2.68E-06 9.99E-08 −7.97E-07 2.68E-06 9.99E-08

9.61E-10 6.87E-11 4.50E-10 3.92E-04 1.58E-05 2.28E-04 2.46E-08 5.39E-07 1.35E-06 1.39E-06 5.39E-07 1.36E-06

< 0.01 < 0.01 < 0.01 0.37 0.25 0.86 < 0.01 < 0.01 < 0.01 0.57 < 0.01 < 0.01

1.30E-01 1.00E-01 9.00E-02 1.00E-03 2.00E-03 7.40E-03 1.10E-01 9.00E-02 8.00E-02 4.00E-03 1.20E-01 9.00E-02

4.25E+02 5.61E+02 4.64E+02 2.32E+02 7.58E+02 1.90E+02 2.56E+03 9.51E+02 4.71E+02 4.59E+02 9.51E+02 4.71E+02

258 258 258 285 285 285 258 258 258 258 258 258

−4.93E-10 −6.21E-11 −7.74E-10 1.54E-04 2.68E-20 −2.22E-05 −5.60E-07 −1.61E-08 −4.78E-07 −1.51E-05 −2.31E-06 −2.74E-05

5.97E-10 4.72E-11 4.25E-10 3.93E-04 3.60E-21 1.54E-05 4.32E-07 1.68E-07 2.72E-07 2.52E-05 1.11E-05 1.66E-05

0.04 0.01 0.02 0.70 0.28 0.17 0.03 0.03 0.02 0.02 0.04 0.03

7.28E-02 6.38E-02 1.02E-01 2.57E-03 3.92E-03 8.63E-03 4.55E-02 6.84E-02 8.26E-02 8.56E-02 7.10E-02 4.67E-02

4.97E+02 6.29E+02 5.15E+02 2.00E+02 1.84E+03 3.13E+01 1.55E+02 2.04E+02 1.79E+02 5.68E+01 1.43E+01 3.52E+01

24 24 24 24 24 24 24 24 24 24 24 24

Fig. 2. The relationship between fishing effort (number of anglers) and harvest of fish per individual angler and per hectare of fishing areas. The left and right column describes fishing areas with lower and higher fishing visit rates, respectively. Each point represents an annual summary of data from an individual fishing reach. 5

Fisheries Research 223 (2020) 105440

R. Lyach

Fig. 3. The relationship between fishing effort (number of anglers) and the harvest of fish per one hectare of fishing areas. The left and right column describes fishing areas with lower and higher fishing visit rates, respectively. Each point represents an annual summary of data from an individual fishing reach.

Fig. 4. The relationship between fishing effort (number of anglers) and the representation of rheophilic fish in the overall fish harvest by biomass. The left and right column describes fishing areas with lower and higher fishing visit rates, respectively. Each point represents an annual summary of data from an individual fishing reach. 6

Fisheries Research 223 (2020) 105440

R. Lyach

Table 2 Composition of Cormorant diet over winters 2014/2015 and 2015/2016. Note: %n, percentage of fish prey; %b, percentage of prey biomass. Species

%n

%b

average length [cm]

average weight [g]

length (min-max) [cm]

weight (min-max) [g]

Barbel Barbus barbus Nase Chondrostoma nasus Vimba Bream Vimba vimba Total

1.0 1.1 0.1 2.2

4.4 4.1 0.3 8.8

30 32 25 29

130 310 230 223

9-38 13-46 20-28 9-46

8-470 18-1350 86-320 8-1350

3.6. Recapture rates of stocked fish

populations (Gokcek and Akyurt, 2007; Pegg and Britton, 2011). The significant effect of fishing effort on the percentage of rheophilic fish in the overall harvest may be explained by the higher abundance of other fish species in the ecosystem (mainly roach, bleak, Alburnus Alburnus, European chub, Squalius cephalus, and European perch, Perca fluviatilis). These species could be caught on the hook more often, creating a buffer that protects rheophilic species from being harvested. Anglers could fill their daily bag limit (7 kg of all fish species) with roaches and chubs before encountering a rheophilic fish. It is possible that rheophilic species do not get caught on the hook as long as the fishing effort is low, mostly because intensively stocked fish species (common carp, rainbow trout, Oncorhynchus mykiss, and piscivores) are usually more aggressive when compared to native rheophilic fish (e.g. Sundström et al., 2004).

The recapture rates of stocked fish were not significantly affected by either fishing effort, intensity of fish stocking, or population density of cormorants (Table 1). This study could not reliably explain the recapture rates of stocked barbel, nase, or vimba bream. 3.7. Cormorant diet analysis Overall 1 318 cormorant pellets were collected and 8 262 diagnostic elements were extracted, which accounted for 5 368 individual fish with a total estimated weight of 489.37 kg. Cormorants consumed fish of 8–49 cm in length and 0.5–1 350 g in weight. The average size among all fish consumed was 18 cm (TL, total length in cm) and 87 g. Rheophilic fish species formed less than 10 % of cormorant diet by number and biomass (Table 2). The length range and weight range of rheophilic fish consumed by cormorants were 9−46 cm (TL, total length) and 8-1 350 g, respectively. The diet of cormorants consisted of 24 fish species from 6 families. Roach, Rutilus rutilus, dominated the diet (60 % by number and 50 % by biomass).

4.3. Cormorant predation Cormorants were observed to frequently prey on shoaling fish which makes rheophilic fish species a suitable prey (Keller, 1995; Suter, 1995 and 1997). Cormorants were also observed to regularly prey on smaller fish (< 20 cm and 100 g), similar in size to rheophilic fish that are stocked in the study area. However, mostly larger fish (20–30 cm and 150–300 g) were identified in cormorant diets. Other researchers mostly found that cormorants preyed on significantly smaller rheophilic fish when compared to cormorants in this study (Keller, 1995; Suter, 1995 and 1997). Another study (Skov et al., 2014) described how cormorants prefer to prey upon larger fish when both small (< 10 cm TL) and larger (10–30 cm TL) fish are available. Rheophilic fish were less abundant prey items in the cormorant diet in comparison to roach, which dominated in the diet. Based on fish monitoring performed by other researchers in the study area (Prchalová et al., 2011; Valová et al., 2014), results suggest that cormorants likely preyed on the most abundant and available fish species. There was no obvious correlation between cormorant population density and the harvest of rheophilic fish, presumably because of differences in size selectivity (Troynikov et al., 2013). However, that does not necessarily mean that cormorants have no significant effect on the decreasing populations of rheophilic species in the area (Lusk, 1996). Cormorants are highly mobile birds, and therefore their predation pressure on fish stocks could have an equal impact in all places because cormorants fill up the ecological niche.

4. Discussion 4.1. Fish harvest The fact that harvest per hectare mostly increased with fishing effort suggests that all rheophilic fish species should be at relatively good population densities and not close to population collapse. However, the fact that harvest per effort decreased with an increase in fishing pressure could be due to the increasing popularity of the catch-and-release fishing strategy in central Europe (Lyach and Čech, 2018). It is also possible that the harvest decreased because fish are getting more timid as previous studies have suggested (Arlinghaus et al., 2017). Increased predation pressure from cormorants, together with increased fishing pressure by anglers, could at some point negatively affect fish populations in the study area. With both parameters working simultaneously, resident fish populations could experience population decreases. A previous study by Lyach and Čech (2018) described a generalised decrease in fish harvest within the area in which this study was conducted. However, this decreased fish harvest can be also linked to the increasing popularity of catch-and-release fishing strategies in central Europe (Lyach and Čech, 2018). To properly test this hypothesis, a tag-and-release study would have to be conducted.

4.4. Data limits

4.2. Fish stocking

Statistics from fisheries have a limit in use and therefore should be interpreted with caution. These are the over/underestimation of size and quantity of harvested fish; incorrect identification of species; an inability of anglers to comply with fishing rules; preferences for specific species; and the popularity of catch-and-release fishing (Essig and Holliday, 1991; Pollock et al., 1994; Cooke et al., 2000; Bray and Schramm, 2001; Mosindy and Duffy, 2007). For a more detailed description of deficiencies in these data, see Lyach and Remr (2019). The cormorant census also has a limit in its use. Cormorants are relatively large and mobile birds that can hunt within a radius of 15–25 km around their roosting sites (Keller, 1995; Suter, 1995 and 1997). Birds that were counted on specific fishing areas could have also

No effect of fish stocking on the fish harvest was observed. The recapture rates of stocked fish should be closely correlated to the intensity of fish stocking, yet no such correlation was observed. It is possible that stocked rheophilic fish have a high natural mortality as previous studies have suggested (Gokcek and Akyurt, 2007; Bašić and Britton, 2016). It is also possible that anglers prefer catching common carp and piscivores to rheophilic fish due to their higher quality of meat (Jankovský et al., 2011). Other studies also claim that stocking rheophilic fish has little or no effect on harvest rates unless the fish are stocked at a very high intensity or into rivers with small wild 7

Fisheries Research 223 (2020) 105440

R. Lyach

hunted in a different area later in the day. However, only birds that were present on fishing areas were counted, avoiding double counting cormorants. Smaller streams were excluded from the cormorant census because cormorants usually prefer to hunt on larger rivers (Keller, 1995; Suter, 1995 and 1997) and few birds were observed on smaller streams in the area (Czech Fishing Union, unpubl. data). Since the analysis was based on six days of cormorant census and eight days of collecting pellets, the error introduced in this study may be significant. The variance in cormorant population densities between the studied fishing areas was quite high. Therefore, it is less likely that the weak relationship between cormorant population density and the harvest of rheophilic fish was due to the equal impact of cormorant predation among all fishing areas. Other researchers used otoliths to identify fish species in the diet of cormorants, stating that fish bones are sometimes not preserved in pellets of cormorants (see Barrett et al., 2007 for review). However, we managed to extract preserved and mostly intact fish bones from the collected cormorant pellets. Generalized linear mixed models were used in statistical analyses because the variable “fishing reach” was used as a random effect in the models. We had no information regarding angling preferences of individual anglers in individual fishing areas. Some anglers may fish only in specific fishing areas, while others may travel around the region and fish in multiple fishing areas. This study did not aim to discover differences in angler preferences on individual fishing areas.

potential ecological and managerial implications. Fish Fish. 18 (2), 360–373. https:// doi.org/10.1111/faf.12176. Baird, O.E., Krueger, C.C., Josephson, D.C., 2006. Growth, movement, and catch of brook, rainbow, and brown trout after stocking into a large, marginally suitable Adirondack river. N. Am. J. Fish. Manag. 26 (1), 180–189. https://doi.org/10.1577/M04-183.1. Barrett, R.T., Camphuysen, K., Anker-Nilssen, T., Chardine, J.W., Furness, R.W., Garthe, S., Hüppop, O., Leopold, M.F., Montevecchi, W.A., Veit, R.R., 2007. Diet studies of seabirds: a review and recommendations. ICES J. Mar. Sci. 64 (9), 1675–1691. Bartholomew, A., Bohnsack, J.A., 2005. A review of catch-and-release angling mortality with implications for no-take reserves. Rev. Fish Biol. Fish. 15 (1-2), 129–154. Bašić, T., Britton, J.R., 2016. Characterizing the trophic niches of stocked and resident cyprinid fishes: consistency in partitioning over time, space and body sizes. Ecol. Evol. 6 (14), 5093–5104. https://doi.org/10.1002/ece3.2272. Bates, D., Maechler, M., Bolker, B., Walker, S., 2015. Fitting linear mixed-effects models using lme4. J. Stat. Softw. 67 (1), 1–48. Brinker, A., Chucholl, C., Behrmann‐Godel, J., Matzinger, M., Basen, T., Baer, J., 2018. River damming drives population fragmentation and habitat loss of the threatened Danube streber (Zingel streber): Implications for conservation. Aquatic Conservation: Marine and Freshwater Ecosystems 28 (3), 587–599. Birkeland, C., Birkeland, C., Dayton, P.K., 2015a. The Importance in Fishery Management of Leaving the Big Ones The importance in fishery management of leaving the big ones. Trends Ecol. Evol. (Amst.) 20 (7), 3–6. https://doi.org/10.1016/j.tree.2005.03. 015. Birkeland, C., Birkeland, C., Dayton, P.K., 2015b. The Importance in Fishery Management of Leaving the Big Ones The importance in fishery management of leaving the big ones. Trends Ecol. Evol. 20 (7), 3–6. https://doi.org/10.1016/j.tree.2005.03.015. Boukal, D.S., Jankovský, M., Kubečka, J., Heino, M., 2012. Stock-catch analysis of carp recreational fisheries in Czech reservoirs: insights into fish survival, water body productivity and impact of extreme events. Fish. Res. 119–120, 23–32. https://doi. org/10.1016/j.fishres.2011.12.003. Bray, G.S., Schramm, H.L., 2001. Evaluation of a state-wide volunteer angler diary program for use as a fishery assessment tool. N. Am. J. Fish. Manag. 21, 606–615 DOI: 10.1577/1548-8675(2001)021 < 0606:EOASVA > 2.0.CO;2. Čech, M., Čech, P., 2013. The role of floods in the lives of fish-eating birds: predator loss or benefit? Hydrobiologia 717, 203–211. https://doi.org/10.1007/s10750-0131625-3. Čech, M., Čech, P., 2017. Effect of brood size on food provisioning rate in Common Kingfisher Alcedo atthis. Ardea 105, 5–17. Čech, M., Vejřík, L., 2011a. Winter diet of great cormorant (Phalacrocorax carbo) on the River Vltava: estimate of size and species composition and potential for fish stock losses. Folia Zool. Brno 60, 129–142. Čech, M., Vejřík, L., 2011b. Winter diet of great cormorant (Phalacrocorax carbo) on the River Vltava: estimate of size and species composition and potential for fish stock losses. Folia Zool. Brno 60 (2), 129–143. Čech, M., Čech, P., Kubečka, J., Prchalová, M., Draštík, V., 2008. Size Selectivity in Summer and Winter Diets of Great Cormorant (Phalacrocorax carbo): Does it Reflect Season-Dependent Difference in Foraging Efficiency? Waterbirds 31, 438–447. Cooke, S.J., Dunlop, W.I., McLennan, D.M., Power, G., 2000. Applications and characteristics of angler diary programs in Ontario, Canada. Fish. Manag. Ecol. 7, 473–487. https://doi.org/10.1046/j.1365-2400.2000.00232.x. Dudgeon, D., Arthington, A.H., Gessner, M.O., Kawabata, Z.I., Knowler, D.J., Lévêque, C., Nailman, R.J., Prieur-Richard, A.H., Soto, D., Stiassny, M.L.J., Sullivan, C.A., 2006. Freshwater biodiversity: importance, threats, status and conservation challenges. Biol. Rev. Camb. Philos. Soc. 81 (2), 163–182. https://doi.org/10.1017/ S1464793105006950. Essig, R.J., Holliday, M.C., 1991. Development of a recreational fishing survey: the marine recreational fishery statistics survey case study. Am. Fish. Soc. 12, 245–254. FAO, 2010. The State of World Fisheries and Aquaculture – 2010 (SOFIA). Food and Agriculture Organisation of the United Nations (FAO), Rome, Italy. Florea, L., 2008. 2011. Characteristics of freshwater fish population in the Danube river near Braila within the period 2006-. J. Environ. Protection and Ecol. 12 (3), 814–823. Gokcek, C.K., Akyurt, I., 2007. The effect of stocking density on yield, growth, and feed efficiency of himri barbel (Barbus luteus) nursed in cages. Israeli J. Aquacult.Bammidgeh 59, 99–103. Golski, J., Pińskwar, P., Jezierska-Madziar, M., Andrzejewski, W., Mazurkiewicz, J., Stanisławski, D., Urbańska, M., 2017. Rheophilic fish in oxbow lakes of the Warta River?the effect of environmental conditions on habitat selection. Oceanological and Hydrobiological Studies 46 (1), 38–49. Gwinn, D.C., Allen, M.S., Johnston, F.D., Brown, P., Todd, C.R., Arlinghaus, R., 2015. Rethinking length‐based fisheries regulations: the value of protecting old and large fish with harvest slots. Fish Fish. 16 (2), 259–281. https://doi.org/10.1111/faf. 12053. Hadfield, J.D., 2010. MCMC Methods for Multi-Response Generalized Linear Mixed Models: The MCMCglmm R Package. J. Stat. Softw. 33 (2), 1–22. Humpl, M., Pivnička, K., Jankovský, M., 2009. Sport fishery statistics, water quality, and fish assemblages in the Berounka River in 1975-2005. Folia Zool. Brno 58 (4), 457–465. Jankovský, M., Boukal, D.S., Pivnička, K., Kubečka, J., 2011. Tracing possible drivers of synchronously fluctuating species catches in individual logbook data. Fish. Manag. Ecol. 18 (4), 297–306. https://doi.org/10.1111/j.1365-2400.2011.00783.x. Jepsen, N., Ravn, H.D., Pedersen, S., 2018. Change of foraging behavior of cormorants and the effect on river fish. Hydrobiologia 820 (1), 189–199. Keckeis, H., Winkler, G., Flore, L., Reckendorfer, W., Schiemer, F., 1997. Spatial and seasonal characteristics of 0+ fish nursery habitats of nase, Chondrostoma nasus in the River Danube. Austria. Folia Zool. 46, 133–150. Keller, T., 1995. Food of cormorants Phalacrocorax carbo sinensis wintering in Bavaria,

4.5. Conclusion It seems that fishing pressure is the main driver of harvest rates in rheophilic fish species, mostly because fish stocking and increases in cormorant numbers did not significantly affect fish harvests. The recapture rates of stocked fish were also relatively low in comparison to the existing population abundance. Since fishing pressure in the study area has increased over the last decade, future populations of rheophilic fish could be significantly threatened by recreational fisheries. We suggest that future studies should focus on developing a strategy that would use population modelling to help conserve rheophilic fish species and perhaps decrease fishing pressure in the area. The author states that he has no conflict of interest (conflict of interest: none). Declaration of Competing Interest The author states that he has no conflict of interest (conflict of interest: none). Acknowledgements The Czech Fishing Union (namely Jaroslava Fryšová, Pavel Horáček, and Dušan Hýbner) provided the necessary fishery data. Pavel Vrána, Karel Anders, and Robert Arlinghaus provided helpful insights into recreational fishing. Martin Čech helped with data collection and analyses. Otakar Ďurďa and Marek Omelka provided help regarding the statistical analyses. Anglers and angling guards in the Czech Republic collected data for this study and therefore made this study possible. This study did not receive any funding. Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.fishres.2019.105440. References Arlinghaus, R., Laskowski, K.L., Alós, J., Klefoth, T., Monk, C.T., Nakayama, S., Schröder, A., 2017. Passive gear‐induced timidity syndrome in wild fish populations and its

8

Fisheries Research 223 (2020) 105440

R. Lyach southern Germany. Ardea 83 (1), 185–192. Keller, T., 1998. 1998. The food of cormorants (Phalacrocorax carbo sinensis) in Bavaria. J. fur Ornithologie. Lajus, J., Kraikovski, A., Lajus, D., 2013. Coastal fisheries in the Eastern Baltic Sea (Gulf of Finland) and its basin from the 15 to the early 20th centuries. PLoS One 8 (10). https://doi.org/10.1371/journal.pone.0077059. Lehikoinen, A., Heikinheimo, O., Lehtonen, H., Rusanen, P., 2017. The role of cormorants, fishing effort and temperature on the catches per unit effort of fisheries in Finnish coastal areas. Fish. Res. 190, 175–182. https://doi.org/10.1016/j.fishres. 2017.02.008. Lusk, S., 1996. Development and status of populations of Barbus barbus in the waters of the Czech Republic. Folia Zool. Brno 45, 39–46. Lyach, R., Remr, J., 2019. The effects of environmental factors and fisheries management on recreational catches of perch Perca fluviatilis in the Czech Republic. Aquat. Living Resour. 32, 15. Lyach, R., Čech, M., 2018. A new trend in Central European recreational fishing: more fishing visits but lower yield and catch. Fish. Res. 201, 131–137. https://doi.org/10. 1016/j.fishres.2018.01.020. Lyach, R., Blabolil, P., Čech, M., 2018. Great Cormorants Phalacrocorax carbo feed on larger fish in late winter. Bird Study 65 (2), 249–256. Mueller, M., Pander, J., Geist, J., 2018. Comprehensive analysis of& 30 years of data on stream fish population trends and conservation status in Bavaria, Germany. Biol. Conserv. 226, 311–320. https://doi.org/10.1016/j.biocon.2018.08.006. Michaletz, P.H., Wallendorf, M.J., Nicks, D.M., 2008. Effects of stocking rate, stocking size, and angler catch inequality on exploitation of stocked channel catfish in small Missouri impoundments. N. Am. J. Fish. Manag. 28 (5), 1486–1497. Mosindy, T.E., Duffy, M.J., 2007. The use of angler diary surveys to evaluate long-term changes in muskellunge populations on Lake of the Woods, Ontario. Environ. Biol. Fishes 79, 71–83. Nakagawa, S., Johnson, P.C., Schielzeth, H., 2017. The coefficient of determination R 2 and intra-class correlation coefficient from generalized linear mixed-effects models revisited and expanded. J. R. Soc. Interface 14 (134), 213–217. Pegg, J., Britton, J.R., 2011. Effects of inter-and intra-specific competition on the growth rates of juvenile European barbel Barbus barbus used in the stock enhancement of UK fisheries. Fish. Res. 112 (1-2), 8–12. https://doi.org/10.1016/j.fishres.2011.08.003. Penczak, T., Glowacki, L., Galicka, W., Koszalinski, H., 1998. A long-term study (19851995) of fish populations in the impounded Warta River, Poland. Hydrobiologia 368 (1), 157–173. https://doi.org/10.1023/a:1003246115666. Piria, M., Simonović, P., Zanella, D., Ćaleta, M., Šprem, N., Paunović, M., Tornjanovic, T., Gavrilovic, A., Pecina, M., Spelic, I., Matulic, D., Rezic, A., Anicic, I., Safner, R., Treer, T., 2019. Long-term analysis of fish assemblage structure in the middle section of the

Sava River – The impact of pollution, flood protection and dam construction. Sci. Total Environ. 651, 143–153. https://doi.org/10.1016/j.scitotenv.2018.09.149. Pollock, K.H., Jones, C.M., Brown, T.L., 1994. Angler Survey Methods and Their Applications in Fisheries Management. American Fisheries Society Special Publication 25, Bethesda, pp. 371 pp. Prchalová, M., Horký, P., SlavíK, O., Vetešník, L., Halačka, K., 2011. Fish occurrence in the fishpass on the lowland section of the River Elbe, Czech Republic, with respect to water temperature, water flow and fish size. Folia Zool. Brno 60 (2), 104–114. Schletterer, M., Kuzovlev, V.V., Zhenikov, Y.N., Tuhtan, J.A., Haidvogl, G., Friedrich, T., Górski, K., Füreder, L., 2018. Fish fauna and fisheries of large European rivers: examples from the Volga and the Danube. Hydrobiologia 814 (1), 45–60. Skov, C., Jepsen, N., Baktoft, H., Jansen, T., Pedersen, S., Koed, A., 2014. Cormorant predation on PIT-tagged lake fish. J. Limnol. 73 (1). Sundström, L.F., Petersson, E., Höjesjö, J., Johnsson, J.I., Järvi, T., 2004. Hatchery selection promotes boldness in newly hatched brown trout (Salmo trutta): implications for dominance. Behav. Ecol. 15 (2), 192–198. Suter, W., 1995. Are Cormorants Phalacrocorax carbo wintering in Switzerland approaching carrying capacity? An analysis of increase patterns and habitat choice. Ardea 83 (1), 255–266. Suter, W., 1997. Roach rules: shoaling fish are a constant factor in the diet of cormorants Phalacrocorax carbo in Switzerland. Ardea 85 (1), 9–27. Takai, N., Kawabe, K., Togura, K., Kawasaki, K., Kuwae, T., 2018. The seasonal trophic link between Great Cormorant Phalacrocorax carbo and ayu Plecoglossus altivelis altivelis reared for mass release. Ecol. Res. 33 (5), 935–948. Tonolla, D., Bruder, A., Schweizer, S., 2017. Evaluation of mitigation measures to reduce hydropeaking impacts on river ecosystems–a case study from the Swiss Alps. Sci. Total Environ. 574, 594–604. https://doi.org/10.1016/j.scitotenv.2016.09.101. Trigo, F.A., Roberts, C.G., Britton, J.R., 2017. Spatial variability in the growth of invasive European barbel Barbus barbus in the River Severn basin, revealed using anglers as citizen scientists. Knowl Manag Aquat Eco 418, 17. Troynikov, V., Whitten, A., Gorfine, H., Pūtys, Ž., Jakubavičiūtė, E., Ložys, L., Dainys, J., 2013. Cormorant catch concerns for fishers: estimating the size-selectivity of a piscivorous bird. PLoS One 8 (11), e77518. Valová, Z., Janáč, M., Švanyga, J., Jurajda, P., 2014. Structure of 0+ juvenile fish assemblages in the modified upper stretch of the River Elbe, Czech Republic. Czech J. Anim. Sci. 59 (1), 35–44. Wiley, R.W., Whaley, R.A., Satake, J.B., Fowden, M., 1993. Assessment of stocking hatchery trout: a Wyoming perspective. N. Am. J. Fish. Manag. 13 (1), 160–170. Yule, D.L., Whaley, R.A., Mavrakis, P.H., Miller, D.D., Flickinger, S.A., 2000. Use of strain, season of stocking, and size at stocking to improve fisheries for rainbow trout in reservoirs with walleyes. N. Am. J. Fish. Manag. 20 (1), 10–18.

9