Abstract
Small lakes provide substantial ecosystem services to society, particularly recreational services. These ecosystem services are rarely quantified. It is also unclear whether expectations about desired lake attributes by various user groups and the public at large align. In many landscapes most small lakes artificially originate from sand and gravel mining along highways and close to cities. Using a choice experiment, in samples from north-western Germany we quantified the ecological services provided by these so-called gravel pit lakes and investigated potential conflicts among the preferences of the public and anglers as a specific user group. The same visual choice sets were employed to assess the preferences for different recreational services as well as biodiversity attributes. We also tested whether the presence of anglers at the small gravel pit lakes affected how people from the public valued the services provided by the lakes. Both the public and the anglers valued the presence of endangered fish species and endangered other taxa positively, suggesting that improved conservation of biodiversity would benefit all users. The public and anglers varied in relation to the recreational uses desired at gravel pit lakes. The opportunity to swim was valued positively by the public and negatively by the surveyed anglers. Yet, the presence of anglers did not significantly reduce the value assigned to gravel pit lakes by the public, suggesting the co-existence of anglers and swimmers would be possible. Co-existence could be fostered through establishing separate angling and swimming zones. Our work suggests the presence of local trade-offs in the management of small gravel pit lakes, specifically between swimmers and anglers. However, conservation of endangered species constitutes a common goal that suits the expectations of all.
Introduction
Lakes are a widespread element in many landscapes. Most lakes worldwide are small with an area between one and ten hectares (Cael et al., 2017). This applies particularly to artificial lakes, such as ponds or gravel pit lakes, which are common components of most urbanized landscapes (Søndergaard et al., 2018). Lakes in general provide several ecosystem services to society (Reynaud and Lanzanova, 2017), but there is limited quantification of the social value of small artificially created lakes (Meyerhoff et al., 2019). In Germany, many gravel pit lakes are managed by recreational angling clubs (Matern et al., 2019), and fishing rights and management duties for fish are tied to ownership of lakes (Daedlow et al., 2011). Additional public use of lakes for outdoor recreation, such as swimming and walking, without specific permission of the fishing rights holder is common. This means that different outdoor recreationists overlap in many localities, which can cause conflicts (Meyerhoff et al., 2019).
Policy decisions and management actions by landowners or fishing rights holders can affect the quality of local lake ecosystems. For example, if a lake is situated in a conservation area, authorities may constrain the use and access of certain user groups, such as anglers, with the objective to reduce disturbances of wildlife or habitats (Arlinghaus, 2006). As another example, agriculture and water management decisions will affect nutrient flows and can affect water visibility and habitat quality for endangered species or biodiversity (Davies et al., 2008) or reduce recreational values. To inform fishing and non-fishing related policy choices, the values assigned to elements of lakes by both the public and specific user groups, such as anglers, are important to know.
The objective of this work was to elicit the preferences of both the public and anglers for attributes of gravel pit lake ecosystems. Using visual choice experiments (Meyerhoff et al., 2019), we examined preferences for ecological and habitat-structural attributes to understand possible trade-offs in the recreation-related ecosystem services generated by gravel pit lakes as perceived by the public and anglers as a distinct user group with management sovereignty in Germany. Additionally, we tested for a potential impact of angling on the public's valuation of the ecosystem services provided by gravel pits. Addressing this question is relevant because of increasing policy interventions directed at constraining access by recreational anglers to rivers and lakes in Germany, motivated with reference to European and national conservation legislation assuming that anglers disturb habitats and wildlife. This is particularly true for conservation efforts in the context of the EU - Habitats Directive (Council Directive 92/43/EEC), the connected EU - Natura 2000 network, and the EU Biodiversity Strategy 2030 as part of the European Green Deal where 30 % of the land and 30 % of the sea within all member states should be protected landscapes in the future (European Commission, 2020). We thus hypothesized that the presence of recreational anglers would reduce the value assigned by the public to gravel pit lakes, perhaps because anglers might be perceived as interfering with other desired recreational activities or as a threat to biodiversity.
Method
Choice experiments (CE) are a preference assessment technique that asks individuals to select their preference among mutually exclusive alternatives (Holmes et al., 2017). CE have become popular among survey-based techniques to elicit individual preferences for changes in the provision of environmental goods and services. As output, CE provide a wide range of information for decision-making and management. For example, the choices recorded enable estimates of the trade-offs among attribute levels individuals are willing to make, giving insights into their preferences. If one of the attributes is a cost variable, marginal willingness to pay (WTP) estimates can be calculated, representing peoples’ preferences for different attributes on the same (monetary) unit (for an application to lakes, see Meyerhoff et al., 2019). These data can then be incorporated into cost-benefit analyses or for compensation payments. We applied a visual choice task where attributes and levels were presented in generated images (Figure 1).
Our attributes and levels (Table 1) in the CE were chosen based on the results of pre-test workshops with angling clubs, reflecting the attributes anglers considered of greatest importance to their quality assessments of gravel pit lakes (Meyerhoff et al., 2019). Two attributes captured the abundance of prey (“non-game”) and predatory (“game”) fish. A higher abundance would change the probabilities to catch fish and is thus an important characteristic for many anglers. Another attribute aimed at the percentage of the lake's shoreline that would be easily accessible for recreation. It was operationalized by the amount of woody vegetation at the shorelines relative to total shoreline length. Accessibility is supposed to be an important attribute for both the public and anglers as it impacts the ease with which some recreational activities can be carried out. Two additional attributes captured whether endangered species (fish and other taxa) would be present at and in the lake. We included this attribute presuming that people would value the presence of endangered species as indicative of biodiversity, even if their level of local biodiversity knowledge can be considered low for the public. The assumption that people would still value the presence of endangered species relates to the general high pro-ecological belief system in contemporary German culture that has recently been confirmed through a national nature awareness study (Nürnberg and Edrmann 2019). Regarding recreational activities, we added attributes indicating the opportunity of walking, swimming, or boating at the lake in addition to fishing. The final attribute was a payment that the individual would have to make to achieve the chosen outcomes in terms of attribute levels.
Attribute . | Levels . |
---|---|
Endangered species (other than fish) | yes / no |
Endangered fish species | yes / no |
Prey fish | low (8), medium (16), high (24) |
Predatory fish | low (4), medium (8), high (16) |
Accessibility of lake shores | 10% / 30% / 70% / 90% |
Walking | yes / no |
Swimming | yes / no |
Boating (angling or pleasure) | yes / no |
General public: Tax to development fund in € per year | 5 / 20 / 50 / 90 |
Angler: Increased club fee in € per year | 5 / 20 / 50 / 90 |
Attribute . | Levels . |
---|---|
Endangered species (other than fish) | yes / no |
Endangered fish species | yes / no |
Prey fish | low (8), medium (16), high (24) |
Predatory fish | low (4), medium (8), high (16) |
Accessibility of lake shores | 10% / 30% / 70% / 90% |
Walking | yes / no |
Swimming | yes / no |
Boating (angling or pleasure) | yes / no |
General public: Tax to development fund in € per year | 5 / 20 / 50 / 90 |
Angler: Increased club fee in € per year | 5 / 20 / 50 / 90 |
Note: For prey fish and predatory fish, the numbers in parenthesis give the number of fish each time presented on the visual choice sets.
Attribute levels were allocated to alternatives via an experimental design optimised for a multinomial logit model using the D-efficiency criterion. To inform the design, we used weak priors, i.e. narrow intervals close to zero from a uniform random distribution to indicate expected parameter signs. A constraint for the design was that the abundance of prey fish had to be larger or equal to that of predatory fish, following classical aquatic food webs where prey is more numerous than predators and energy is lost from lower trophic levels to the upper ones (Elton, 1927). Note that fish abundance of selected species in a lake can be also increased by stocking and does not only rely on natural reproduction, which is why it is plausible to independently change either prey or predatory fishes. The final design comprised 32 choice sets that were blocked into four subsets of each eight sets. The exact same experimental design was used in all surveys. Instead of presenting choice sets in a common text-based format, we generated visual choice sets (Figure 1; Meyerhoff et al., 2019).
Prior to presenting the valuation scenarios, respondents were introduced to the choice attributes. A short definition together with an icon representing the attribute was given (Fig. 1), and respondents were asked to memorise the icon. Moreover, those among the public as well as anglers who had visited a lake within 12 months prior to the survey were requested to indicate the current situation at the last visited lake or their homelake. Homelakes were defined as the lake people were most familiar with. Attribute levels were displayed to respondents, and they were requested to select the one that best matched the current situation.
Although we aimed at keeping the valuation context as constant as possible across the surveys, scenarios had to differ in some respects. In the scenario provided to the public, respondents were asked to imagine that the government of the Federal State of Lower Saxony (Germany) plans to implement a programme to increase the quality of small lakes. Depending on whether respondents had visited a lake during the year prior to the survey, they were informed that the improvements would be carried out at the lake of their last visit or that a decision about which lakes to improve had not been made. In either case, they were told that their preferences would influence the design of the enhancement programme. The payment requested for the alternatives was said to finance the corresponding quality improvements preferred by people. Additionally, respondents were informed that this money would not be available for other expenses and that they could also choose the option that no programme designed to improve the quality of lakes would be implemented.
Because angling might be controversial to other recreationists and interfere with other recreational activities, such as swimming, we experimentally tested whether angling at the lakes affected preferences by the public. In the public survey, we used two treatments allowing recreational fishing (“With fishing”) or not (“Without fishing”). Accordingly, for the latter treatment the angler visible at the shore of the lake in Figure 1 was removed for the treatment “Without fishing”. Respondents, who were randomly assigned to one of the treatments, were additionally informed whether fishing was allowed or not.
In the scenarios provided to anglers, respondents were asked to imagine that their club could add a new gravel pit lake to the current portfolio of water bodies for angling, and the club must decide how to manage it. Respondents were then told that the objective of the survey is to elicit their preferences regarding this new lake. The new lake could differ in the attributes used to record the perceived quality of their home lake, i.e. the lake they are most familiar with at their club but would have the same size as the lake and would be just as far away to keep these characteristics constant. Another difference was that the payment for anglers was added to current club fees, and thus the text accompanying the coin indicating the amount of money respondents had to spent was different to the public survey. In the public version, it said “Payment to the fund”, and in the angler version it said, “Higher club fee”. Each survey was run by a different survey company, and a different survey mode was used. The public survey used an online panel to recruit respondents from the population of Lower Saxony. The angler survey used a mail survey to invite randomly selected anglers from the membership inventories of ten angling clubs in Lower Saxony. Anglers also had the opportunity to answer the survey online.
To conduct the interviews with people from the public in Lower Saxony, a random sample from the online panel of the survey company Respondi, Colone, was drawn. The sample aimed to be representative for people in Lower Saxony above 18 years using quotas for gender and age. People were invited to participate by email and carried out the survey online; it took place in June and July 2018. To conduct the interviews among anglers in Lower Saxony, respondents were randomly drawn from member lists of 10 angling clubs. All selected anglers were invited by mail, providing them with a printed questionnaire and at the same time offering a link to respond online. Each angler was invited with a letterhead of his or her angling club. Overall, 5.500 members across the ten clubs received the invitation to participate. The survey implementation followed the approach suggested by Dillman (2000). This survey took place in late 2017 and early 2018.
Modelling approach
where ωn = βn/αn, that is, WTP for non-price attributes x. The price attribute parameter α is assumed to follow a lognormal distribution, the WTP parameters ω are assumed to follow a normal distribution.
In both surveys, respondents were asked to indicate the situation at the lake they had visited last with respect to the attribute levels. The stated values for the perceived quality were then assigned to the status quo alternative as in Meyerhoff et al. (2019). In case respondents had not visited any lake in the 12 months prior to the interview or had not stated values for all attributes, the values were imputed by mean values or the median calculated from those who had visited and responded to this question. The same procedure was used for the anglers except that the imputation was done on the level of the homelake. This way, the imputed value was lake-specific and not an overall mean value as in the public survey. Effects not captured by the attributes were taken by two alternative specific constants (ASC), one for the first alternative displayed on the left of the visual choice set (ASCfirst) and a second for the status quo alternative (ASCsq; No improvement of gravel pits / staying with the homelake). To investigate the effects of the treatments angling allowed / not allowed, we generated interaction effects between all attributes and a dummy variable indicating if the respondent was assigned to the treatment “Without fishing”. A statistically significant interaction term would indicate that the mean WTP estimate differs between treatments (fishing allowed vs. not allowed). The cost coefficients were all statistically significant and had the expected negative sign, but as we focus on the WTP values here, we do not report those coefficients here. All models were estimated in Stata using the “mixlogitwtp” command provided by Hole and Kolstad (2012) using 2000 Halton draws to perform the simulation of the log-likelihood.
Results
The public - 648 respondents in the treatment (“With fishing”) and 662 in the treatment “Without fishing” - and the angler sample (1.553 respondents) differed significantly in terms of demographics. The public was on average six years younger, and the share of females was much larger (> 50 %) than in the angler sample (4 %, Table 2). Respondents from the public visited a lake within 12 months prior to the interview on average 6 to 7 times whereas anglers reported almost double amounts of fishing days, i.e. days they had visited their homelake.
. | . | Public . | Angler . | |
---|---|---|---|---|
. | . | With fishing . | Without fishing . | . |
Age (mean/SD) | 44.3 (15.3) | 44.4 (15.0) | 50.2 (15.6) | |
Gender (% female) | 50.0 | 51.2 | 4.0 | |
Highest education level | ||||
Lower secondary school (% yes) | 10.5 | 12.4 | 26.1 | |
Secondary school (% yes) | 39.4 | 33.5 | 33.6 | |
High school (% yes) | 36.1 | 36.6 | 14.7 | |
Lake visit last 12 months (% yes) | 59.2 | 56.2 | 100.0 | |
Lake visits (mean/SD) | 7.12 (19.7) | 5.7 (14.8) | ||
Fishing days (mean/SD) | 14.9 (20.8) | |||
Main activity last visit (%) | ||||
Walking | 34.9 | 37.6 | ||
Swimming / bathing | 20.6 | 21.2 | ||
Cycling | 9.6 | 8.1 | ||
Dog walking | 10.7 | 6.9 | ||
Fishing | 1.8 | 2.7 | 100.0 | |
Respondents | 648 | 662 | 1553 |
. | . | Public . | Angler . | |
---|---|---|---|---|
. | . | With fishing . | Without fishing . | . |
Age (mean/SD) | 44.3 (15.3) | 44.4 (15.0) | 50.2 (15.6) | |
Gender (% female) | 50.0 | 51.2 | 4.0 | |
Highest education level | ||||
Lower secondary school (% yes) | 10.5 | 12.4 | 26.1 | |
Secondary school (% yes) | 39.4 | 33.5 | 33.6 | |
High school (% yes) | 36.1 | 36.6 | 14.7 | |
Lake visit last 12 months (% yes) | 59.2 | 56.2 | 100.0 | |
Lake visits (mean/SD) | 7.12 (19.7) | 5.7 (14.8) | ||
Fishing days (mean/SD) | 14.9 (20.8) | |||
Main activity last visit (%) | ||||
Walking | 34.9 | 37.6 | ||
Swimming / bathing | 20.6 | 21.2 | ||
Cycling | 9.6 | 8.1 | ||
Dog walking | 10.7 | 6.9 | ||
Fishing | 1.8 | 2.7 | 100.0 | |
Respondents | 648 | 662 | 1553 |
Note: Due to missing values age is only calculated based on 1,503 anglers; education levels do not add up to 100% as we only report the major graduation levels.
The share of people among the public who indicated fishing as the main activity during their last lake visit was 1.8 % for the treatment “With fishing” and 2.7 % for the treatment “Without fishing”. Among the remaining recreational activities walking, especially when joined with dog-walking, was very popular among the public (Treatment “With fishing”: 46.6 %; Treatment “Without fishing”: 44.5 %, Table 2). Swimming also revealed a high share among the stated activities during the last lake visit (Treatment “With fishing”: 20.6 %; Treatment “Without fishing”: 21.2 %). We did not ask for other activities than fishing in the survey among anglers. The perception of the status quo at the lake visited or the homelake, respectively, differed between the public and anglers, where lakes described by anglers contained more fish, more endangered species, and less recreational activities other than angling (Table 3).
. | With fishing . | Without fishing . | Angler . | |||
---|---|---|---|---|---|---|
% don't | mean | % don't | mean | % don't | mean | |
Prey fish | 45.6 | 15.4 | 42.5 | 15.0 | 6.5 | 17.2 |
Predatory fish | 44.3 | 5.6 | 43.8 | 6.5 | 7.5 | 7.9 |
Access shore | 6.0 | 53.3 | 9.7 | 55.8 | 7.9 | 51.1 |
End. fish species | 45.3 | 0.1 | 42.2 | 0.1 | 54.8 | 0.2 |
End. other species | 41.9 | 0.3 | 40.9 | 0.3 | 27.1 | 0.7 |
Walking | 5.7 | 0.9 | 4.0 | 1.0 | 4.0 | 0.8 |
Swimming | 9.9 | 0.8 | 7.5 | 0.7 | 5.0 | 0.3 |
Boating | 12.0 | 0.5 | 12.4 | 0.5 | 6.1 | 0.2 |
. | With fishing . | Without fishing . | Angler . | |||
---|---|---|---|---|---|---|
% don't | mean | % don't | mean | % don't | mean | |
Prey fish | 45.6 | 15.4 | 42.5 | 15.0 | 6.5 | 17.2 |
Predatory fish | 44.3 | 5.6 | 43.8 | 6.5 | 7.5 | 7.9 |
Access shore | 6.0 | 53.3 | 9.7 | 55.8 | 7.9 | 51.1 |
End. fish species | 45.3 | 0.1 | 42.2 | 0.1 | 54.8 | 0.2 |
End. other species | 41.9 | 0.3 | 40.9 | 0.3 | 27.1 | 0.7 |
Walking | 5.7 | 0.9 | 4.0 | 1.0 | 4.0 | 0.8 |
Swimming | 9.9 | 0.8 | 7.5 | 0.7 | 5.0 | 0.3 |
Boating | 12.0 | 0.5 | 12.4 | 0.5 | 6.1 | 0.2 |
Note: “% don't” refers to the share of people who stated that they do not know the status quo; mean values refer to those who have responded.
The preferences for lake management differed strongly for most of the attributes in the choice experiment (Table 4 and Figure 2). Starting with the public and the treatment “With fishing”, neither the abundance of prey and predatory fish nor increasing access to the lake shore as main effects significantly impacted respondents’ choices. Thus, having more fish in lakes and more access to the shoreline did not produce value to the public, all else being equal. This was different regarding both types of endangered species. Here, both parameters were highly significant, indicating a WTP of around €28 per year for the presence of endangered fish species and an even higher WTP of €36 per year for the presence of other endangered species. Among recreational activities that could be carried out at the lakes, walking was the most preferred activity by the public. To walk in the surroundings of the lakes, people from the public would be willing to pay around €68 per year whereas for swimming, which was the second important recreational activity, the public would be willing to pay an average of €28 per year. Finally, whether boating was possible at a lake or not was on average of no relevance for the respondents from the public. The respondents did not equally prefer lake characteristics as revealed by highly significant standard deviations of the attribute parameters (lower part Table 3.), indicating preference heterogeneity for all but prey fish. Among the effects of the treatment “Without fishing” only one interaction term indicated a significant difference between both treatments: when angling was not allowed, respondents revealed a preference for higher levels of prey fish. All other lake attributes were not affected by treatment, and thus the WTP values of the public were largely independent of whether angling was allowed or not.
. | Public . | Angler . | |||||||
---|---|---|---|---|---|---|---|---|---|
. | With fishing . | Without fishing . | |||||||
WTP | ∣t-value∣ | 95% CI | WTP | ∣t-value∣ | 95% CI | WTP | ∣t-value∣ | 95% CI | |
Prey fish | 0.23 | 0.80 | -0.34 / 0.81 | 0.96 | 2.15 | 0.08 / 1.83 | 0.54 | 3.17 | 0.21 / 0.89 |
Predatory fish | 0.83 | 1.82 | -.06/ 1.17 | -0.60 | 0.92 | -1.88 / 0.68 | 1.92 | 6.78 | 1.36 / 2.47 |
Access shore | -0.07 | 0.97 | -.22 / 0.07 | 0.05 | 0.46 | -0.15 / 0.25 | -0.07 | 1.78 | -0.15 / 0.01 |
End. fish species | 28.05 | 6.60 | 19.72 / 36.38 | -4.01 | 0.72 | -14.86 / 6.84 | 22.09 | 8.41 | 16.9 / 27.2 |
End. other species | 35.74 | 7.54 | 26.44 / 45.03 | 0.29 | 0.05 | -12.88 / 12.31 | 15.08 | 6.03 | 10.2 / 19.9 |
Walking | 67.62 | 9.72 | 53.99 / 81.25 | 7.13 | 0.09 | -14.86 / 16.28 | -4.17 | 1.63 | -9.19 / 0.86 |
Swimming | 28.16 | 4.41 | 15.65 / 40.67 | 9.20 | 0.99 | -8.93 / 27.34 | -91.40 | 15.74 | -102.7 / −79.9 |
Boating | -1.15 | 0.22 | -11.48 / 9.17 | -4.83 | 0.68 | -18.71 / 9.05 | -22.35 | 5.28 | -30.64 / −14.06 |
ASCfirst | 1.02 | 0.26 | -6.81 / 8.86 | 0.62 | 0.11 | -10.50 / 11.75 | 2.73 | 1.15 | -1.90 / 7.35 |
ASCsq | -144.66 | 8.05 | -179.88 / --109.44 | 22.67 | 1.15 | -16.13 / 61.47 | 21.52 | 3.98 | 10.93 / 32.12 |
Standard deviations | |||||||||
Prey fish | 0.14 | 0.22 | 0.61 | 1.12 | |||||
Predatory fish | 3.03 | 3.97 | 4.69 | 9.77 | |||||
Access shore | 0.85 | 8.53 | 0.61 | 8.31 | |||||
End. fish species | 25.35 | 3.97 | 27.74 | 5.23 | |||||
End. other species | 35.88 | 5.28 | 25.99 | 3.59 | |||||
Walking | 88.79 | 10.18 | 35.97 | 7.34 | |||||
Swimming | 118.95 | 10.87 | 80.56 | 14.69 | |||||
Boating | 78.16 | 10.26 | 100.02 | 15.77 | |||||
ASCfirst | 32.82 | 5.47 | 26.23 | 4.71 | |||||
ASCsq | 242.31 | 10.94 | 139.33 | 16.87 | |||||
Observations | 31,464 | 35,238 |
. | Public . | Angler . | |||||||
---|---|---|---|---|---|---|---|---|---|
. | With fishing . | Without fishing . | |||||||
WTP | ∣t-value∣ | 95% CI | WTP | ∣t-value∣ | 95% CI | WTP | ∣t-value∣ | 95% CI | |
Prey fish | 0.23 | 0.80 | -0.34 / 0.81 | 0.96 | 2.15 | 0.08 / 1.83 | 0.54 | 3.17 | 0.21 / 0.89 |
Predatory fish | 0.83 | 1.82 | -.06/ 1.17 | -0.60 | 0.92 | -1.88 / 0.68 | 1.92 | 6.78 | 1.36 / 2.47 |
Access shore | -0.07 | 0.97 | -.22 / 0.07 | 0.05 | 0.46 | -0.15 / 0.25 | -0.07 | 1.78 | -0.15 / 0.01 |
End. fish species | 28.05 | 6.60 | 19.72 / 36.38 | -4.01 | 0.72 | -14.86 / 6.84 | 22.09 | 8.41 | 16.9 / 27.2 |
End. other species | 35.74 | 7.54 | 26.44 / 45.03 | 0.29 | 0.05 | -12.88 / 12.31 | 15.08 | 6.03 | 10.2 / 19.9 |
Walking | 67.62 | 9.72 | 53.99 / 81.25 | 7.13 | 0.09 | -14.86 / 16.28 | -4.17 | 1.63 | -9.19 / 0.86 |
Swimming | 28.16 | 4.41 | 15.65 / 40.67 | 9.20 | 0.99 | -8.93 / 27.34 | -91.40 | 15.74 | -102.7 / −79.9 |
Boating | -1.15 | 0.22 | -11.48 / 9.17 | -4.83 | 0.68 | -18.71 / 9.05 | -22.35 | 5.28 | -30.64 / −14.06 |
ASCfirst | 1.02 | 0.26 | -6.81 / 8.86 | 0.62 | 0.11 | -10.50 / 11.75 | 2.73 | 1.15 | -1.90 / 7.35 |
ASCsq | -144.66 | 8.05 | -179.88 / --109.44 | 22.67 | 1.15 | -16.13 / 61.47 | 21.52 | 3.98 | 10.93 / 32.12 |
Standard deviations | |||||||||
Prey fish | 0.14 | 0.22 | 0.61 | 1.12 | |||||
Predatory fish | 3.03 | 3.97 | 4.69 | 9.77 | |||||
Access shore | 0.85 | 8.53 | 0.61 | 8.31 | |||||
End. fish species | 25.35 | 3.97 | 27.74 | 5.23 | |||||
End. other species | 35.88 | 5.28 | 25.99 | 3.59 | |||||
Walking | 88.79 | 10.18 | 35.97 | 7.34 | |||||
Swimming | 118.95 | 10.87 | 80.56 | 14.69 | |||||
Boating | 78.16 | 10.26 | 100.02 | 15.77 | |||||
ASCfirst | 32.82 | 5.47 | 26.23 | 4.71 | |||||
ASCsq | 242.31 | 10.94 | 139.33 | 16.87 | |||||
Observations | 31,464 | 35,238 |
Anglers cared substantially about the abundance of both prey and predatory fish, but the WTP was higher for the abundance of predatory fish (€1.9 per abundance unit) than for the abundance of prey fish (€0.5 per abundance unit). Like the public, anglers preferred the presence of both endangered fish and other species, and, similar to the public, the access to the shore attribute was not significant. The WTP of anglers was larger for the presence of endangered fish species (€22.1) than for endangered other species (€15.1) – a pattern different in ranking to the public who had a higher WTP for “other species” than fish. The largest differences between anglers and the public occurred with respect to recreational activities. On average, anglers did not care about other people walking in the vicinity of the lakes, but swimming was valued highly negatively, leading to a mean WTP of €-91.4 per year. Thus, simultaneous swimming and angling activities at the same water body would create a large loss in the utility offered by the lake to anglers. The importance of this attribute for anglers was also underlined by the high t-value (15.9). In addition, boating was valued on average negatively with a loss in benefits of €-22.4 per year. In summary, the average angler disliked lakes with swimmers and boating activities. Like the public, highly significant standard deviations indicated considerable preference heterogeneity among anglers.
Discussion and conclusions
Small gravel pit lakes provide substantial value to both the general public and to anglers. For many of the characteristics of the lakes we investigated, preferences of the public and anglers aligned. Both groups preferred high biodiversity at the lakes, but the public placed higher mean WTP value on both types of endangered species than anglers. Previous survey work revealed that the German public has a rather low connection and knowledge about freshwater fish relative to other European countries (Kochalski et al., 2019, Riepe et al., 2021), which might explain why the values attached to “other species” than fish were higher than those for fishes.
Preferences among the public and anglers differed substantially with respect to recreational activities. Particularly swimming created conflict as indicated by strongly opposing WTP`s among anglers and the public. Recreational activities like swimming compete with anglers for space on small lakes and might disturb angling activities, ultimately reducing the quality of an individual angling experience by creating goal interference (Jacob and Schreyer, 1980). However, the strong average aversion of swimming by anglers as measured in our study might be driven by a small fraction of anglers with very strong aversion of other recreationists, which can disproportionately influence average WTP values (Meyerhoff et al., 2019). Although our data suggest the conflict between swimming and angling activities is among the strongest between different user groups at small lakes, value perceptions of both groups might strongly differ on a local level, depending on the composition of local stakeholders. Moreover, our data show patterns of asymmetric antipathy because the overall value of lakes as perceived by the public was increased when swimming was allowed and not reduced when anglers were present, while swimmers reduced the value of gravel pit lakes as perceived by anglers.
While walking was highly valued by the public, anglers experienced no significant loss of benefits when walking would happen at the lake. Roughly, the positive benefits of three individuals swimming equalled the loss of benefits a single angler, on average, would experience. If in conflict, i.e. both public and anglers have to use the same lake, the balance will mainly depend on how many people from the public prefer to go swimming and how many anglers show up. Deciding which recreational activity should be preferred from an economic point of view will thus be site-specific, and it remains a political question how to weigh the benefits of different user groups. Thus, based on our work alone, whether the benefits derived from recreational activities at a small lake by one user group outweigh the loss in benefits by the other user group cannot be determined in general. This is particularly true for smaller groups of recreationists that can highly differ in their individual preferences. Depending on the local composition of the stakeholders in terms of preferences, at one lake a friendly coexistence might happen between anglers and other outdoor recreationists from the public while at other small lakes preferences for all attributes might be strongly conflicting. In lakes under management of recreational anglers, local experience shows that they attempt to control swimming and other uses through signatures and shoreline management. But from a societal perspective, also swimming creates benefits. Thus, local level benefits supplied by small lakes might be maximised by using different management strategies that range from bans, over spatio-temporal zoning to no action or inclusion of different local lakes in management strategies where values for specific user groups are maximised at different water bodies nearby.
Our research suggests that integrative management of outdoor recreation, particularly angling and swimming, might be necessary under certain local conditions. Where recreational anglers are managers of gravel pits, their action space typically relates to the management of fish stocks and small-scale shoreline developments like free cutting of angling spots, while land and water owners (if different to angling clubs) as well as public agencies typically decide about infrastructure developments in a more general fashion and management of recreational activities, e.g. through protected area ordinances. If anglers are the owner of the water body and the surrounding land, their scope for action increases and often they try to minimize other recreationists along the water body. However, in many cases management concepts on a local level that integrate all common recreational activities do not exist. Spatio-temporal closures or segregations of different recreational activities might be one approach to avoid recreational conflicts and result in co-existence. As a further implication, actions that help endangered species recover can create values that benefit both the public at large and recreational anglers. Recreational angling clubs may thus increasingly rely on habitat enhancement activities because they may help different taxa, including fish to benefit from local conservation actions. Another solution of conflicts between anglers and swimmers might be based on active promotion of mutual appreciation and enhanced communication.
Funding
The study was jointly financed by the German Federal Ministry of Education and Research (BMBF) and the German Federal Agency for Nature Conservation (BfN) with funds granted by the German Federal Ministry for the Environment, Nature Conservation and Nuclear Safety (BMU) within the context of the BAGGERSEE project (www.baggersee-forschung.de). Finally, we are grateful to two anonymous reviewers who provided helpful comments to a prior version of the manuscript and would especially like to thank all anglers and fishing clubs that participated in our choice experiments.