Introduction

Freshwater ecosystems fulfil a wide range of ecological functions. They also provide many services for humans, including fresh water supply (e.g. for consumption and irrigation), food supply (e.g. fish), transportation routes, recreation, hydropower and cooling water (Jackson et al. 2001; Aivazidou et al. 2016). Submerged vegetation is considered vital in shallow water ecosystems for performing a number of ecosystem functions that result in enhanced stability of a clear water state in shallow freshwater ecosystems (Carpenter and Lodge 1986; Jeppesen et al. 1997; Van Donk and Van de Bund 2002; Hilt and Gross 2008). Submerged aquatic plants also provide food and habitat for other autotrophs and both vertebrate and invertebrate animal species and can thereby increase freshwater biodiversity (Hargeby et al. 1994; Schriver et al. 1995; Perrow et al. 1999; Mazzeo et al. 2003; Declerck et al. 2005; Bakker et al. 2016). An important target in restoration of degraded aquatic ecosystems is therefore the restoration of stable and diverse aquatic vegetation (Moss et al. 1996; Hilt et al. 2006; Bakker et al. 2013). However, aquatic plants can also be perceived as problematic, in particular when they occur in large numbers (i.e. mass development) and interfere with human use of the freshwater ecosystem (Chambers et al. 1999; Hilt et al. 2006; Zehnsdorf et al. 2015).

Clear water and high availability of nutrients in the sediment are thought to facilitate this nuisance growth of macrophytes (Barko and Smart 1986; Bini et al. 1999; Angelstein et al. 2009; Verhofstad et al. 2017). Next to nutrient effects, macrophytes generally also grow faster when temperature and light irradiance increase during the seasons (Kunii 1984; Barrat-Segretain 2004; Bornette and Puijalon 2011). The annual timing of strong macrophyte development often coincides with high anthropogenic use of aquatic ecosystems, as water recreation peaks during the summer months. Nuisance growth of invasive submerged macrophytes, such as Myriophyllum spicatum L., Elodea nuttallii (Planch.) St. John, Egeria densa Planch. and Cabomba caroliniana Gray, have caused many problems for humans worldwide (Smith and Barko 1990; Di Nino et al. 2005; Wilson et al. 2007; Tamayo and Olden 2014). However, these effects are not limited to invasive species (Anderson 2003; Hilt et al. 2006), as any macrophyte species that forms large stands may cause nuisance to one of the freshwater ecosystem functions or services.

Whereas nuisance growth of submerged macrophytes is regularly reported in both scientific and popular media, it is largely undefined what ‘nuisance’ actually is. What characteristics (e.g. species composition, plant height, plant cover, etc.) does a vegetation have to be a nuisance to humans? To our knowledge, no quantitative method is available to classify whether submerged vegetation creates nuisance for specific ecosystem services or not. In this study, we developed a quantitative approach to identify and classify submerged nuisance aquatic vegetation.

We performed a review of international scientific literature to: (1) identify which problems are caused by nuisance submerged macrophytes and to which ecosystem services these are related; (2) identify macrophyte characteristics that were used to describe the vegetation as ‘nuisance;’ (3) identify which species have been reported to grow to nuisance levels and whether these are native or non-native species. We used an average Dutch lake as a case study to (4) illustrate our framework classifying a vegetation as either ‘nuisance’ or ‘non-nuisance’ using selected plant characteristics and differentiating according to the functions and services the ecosystem fulfils.

Methods

Literature survey

We performed a systematic search of the scientific literature to find the most common problems reported with submerged macrophytes, the vegetation characteristics linked to these problems and which species most often cause these problems. To find relevant scientific peer-reviewed papers, the Web of Science search engine was used with the following query:

Query: “TOPIC: (nuisance OR noxious OR problematic) AND [macrophyte* OR (aquatic AND plant*) OR (aquatic AND weed*)]”.

This search yielded 346 hits on the 10 February 2017 covering a wide range of aquatic systems and geographic regions (see Electronic Supplementary Material 1). Based on the title and abstract, papers were selected for processing if they contained information on problems caused by submerged freshwater nuisance macrophytes and/or criteria that described macrophytes as nuisance and/or which species were considered a nuisance. Papers describing invasive non-native species were only included in the analysis if the plants were considered a nuisance because of the number at which they occurred, not just because of the fact that they were non-native to that specific site. We found 81 papers that met our criteria from waterways and waterbodies located in North and South America, Europe, Africa and Australia, with the USA being the largest source of scientific studies (ESM 1). Several scientific papers cited local ‘grey literature’ next to peer-reviewed papers and this local knowledge is therefore indirectly also included in this review. From the selected paper, we extracted (1) problems mentioned that were caused by submerged macrophytes, (2) macrophyte vegetation characteristics linked to nuisance and (3) nuisance macrophyte species reported.

Case study

Finally, we used the information gathered from the literature review to develop a framework classifying a macrophyte stand as either nuisance or not, depending on local conditions, such as ecosystem service provided and water depth. To define the threshold levels at which submerged macrophytes cause nuisance that our framework requires, the information in the scientific literature was insufficient. For our case study, we therefore gathered additional information concerning the perception of nuisance in the Netherlands from recreational specialists at Alterra (Wageningen University and Research Centre, The Netherlands), The Netherland’s Sport Fishing Association (SVN) and the Dutch Water Sports Industry Association (HISWA). We also included information from a survey among 1269 interviewed Dutch recreationalists (for more detailed information, see Goossen et al. 1997). This survey contained information on how importantly people value water quality and aquatic vegetation for engaging in recreational activities such as fishing, swimming and boating (Goossen et al. 1997). We used a model shallow (2-m) freshwater ecosystem as our case study, because these ecosystems are very common and typical for The Netherlands and many other countries (e.g. Gulati and Van Donk 2002; Sondergaard et al. 2007). The framework can be tailored to specific waterbodies or waterways by surveying local users about their problems with aquatic vegetation and incorporating corresponding threshold levels of plant abundance into the classification scheme. The thresholds mentioned in our case study (see “Results”) are given specifically for the location that is actually used by the community, not for the entire ecosystem. We further assumed homogeneous vegetation distribution in the case study and will discuss spatial heterogeneity and implications for the whole ecosystem in the discussion.

Results

Recorded problems

A wide range of plant-induced problems have been reported depending on the use of the aquatic system at hand (Table 1). The main problems can be categorised as problems for boat traffic, fishing, swimming and hydrodynamics. Additional problems that were mentioned less frequently included problems concerning decreased aesthetics, increased sedimentation and altered nutrient cycles (grouped under the category ‘Other’ in Table 1). Many studies (n = 38) also mentioned problems that can be categorised as problems for ecological functioning, especially with regard to non-native species. Because these problems for nature do not necessarily directly impair the anthropogenic functions described, they are outside of the scope of this article and included in the ‘Other’ category, but we strongly recommend this to be taken up in future research.

Table 1 Summary of the main services provided by the freshwater ecosystem where nuisance aquatic plants cause problems as stated in the scientific literature
Full size table

Characteristics of nuisance submerged vegetation

Vegetation that was considered to be a nuisance was mainly characterised by a high plant growth rate, tall height, high coverage, forming (near) monospecific vegetation and/or high biomass (ESM 3). Plant height (n = 24) and coverage (n = 45) were the most frequently described plant parameters related to nuisance. Most striking was that only 5 of the 78 scientific papers used in this review provided a quantitative measure of nuisance vegetation, namely >50% plant cover, <0.5 m plant canopy depth or >0.5 m plant height under 1–2 m of water depth (Brandrud and Roelofs 1995; Mataraza et al. 1999; David et al. 2006; Richardson 2008; Alwin et al. 2010). In most papers a specific biomass per square metre, coverage or plant height at which the vegetation was considered as a nuisance was lacking. Several authors did not provide a description of nuisance vegetation but only stated that macrophytes were a nuisance because they ‘caused problems’. Together with the description that plants are a nuisance if they are locally very abundant, these descriptions have been categorised as ‘Other’ (ESM 3), as they cannot easily be attributed to any measurable plant characteristic.

Nuisance species reported

A diverse group of 33 different submerged macrophyte species were reported as being a nuisance (Table 2). Most frequently mentioned (≥6 references) were: M. spicatum, Hydrilla verticillata Royle, Ceratophyllum demersum L., E. densa, Lagarosiphon major Moss, Elodea canadensis Michaux and Potamogeton pectinatus L. M. spicatum, which were reported most often as being a nuisance species, namely in 25 papers. Several macrophyte species have been reported to be a nuisance in both their native and introduced range (7), while others have mainly been reported as nuisance in either their native (17) or introduced ranges (8) (Table 2).

Table 2 Submerged macrophytes recorded in scientific literature as nuisance species or as dominant species in nuisance vegetation and whether they were non-native in the location at hand
Full size table

Determining thresholds: case study

Following the literature (Washington et al. 1992; ESM 3), we propose that classifying nuisance for human use can best be done by combining both plant canopy depth and plant cover as factors (Fig. 1a). With plant cover we indicate the per cent surface area occupied by macrophytes. Plant canopy depth is the difference between water depth and plant height, i.e. the thickness of the water column above the vegetation canopy. Submerged macrophytes become a direct nuisance for humans when too many shoots physically touch humans (swimming) or manmade objects (boating and fishing) in both waterbodies and waterways or when they obstruct flow in waterways (impede hydrodynamics). To separate nuisance from non-nuisance vegetation based on the two selected plant parameters, it is essential to first identify at what canopy depth and cover submerged macrophytes will pose a problem for the different types of ecosystem services the aquatic ecosystem may deliver (Murphy 1988). The first step in determining the threshold levels above which submerged macrophytes will be considered a nuisance for specific ecosystem services is thus to identify which services are provided by the ecosystem. To determine these nuisance threshold levels, we suggest surveying the local users about how many plants are perceived to impair the ecosystem service at hand. Because perception is very subjective and can be exaggerated, we suggest incorporating the available scientific knowledge on actual problems caused by submerged plant growth where possible. Additionally, quantitative metrics about the users of the ecosystem services can also help to remove perception bias. With quantitative metrics we mean for example the size and draught of the recreational boats used for the ecosystem service: boating. In the following sections, we will illustrate in more detail how to determine the threshold values for all ecosystem services mentioned, using The Netherlands as an example.

Fig. 1
figure 1

a Case study (water depth = 2 m) classification of nuisance macrophyte vegetation by plant canopy depth and plant cover depending on ecosystem services provided by the aquatic system, as proposed by the authors for sites in the ecosystem where the functions are performed. Letters indicate that macrophytes are classified as nuisance for each anthropogenic function: B1 (small recreational boats); B2 (large recreational boats); F (fishing); S (swimming); H (hydrodynamics, i.e. water flow in this case). For a graphical representation, see c. Darker (red) fills indicate more services are impaired by macrophytes. b Probability of maintaining two important ecosystem services (i.e. clear water and fish populations) in shallow aquatic systems considering the total area of the ecosystem. Lighter fills are generally considered more desirable for the stability of the clear water state. The school of fish indicates the plant cover that is suggested as optimal for fish populations. See “Discussion”: ‘Aquatic plant problems versus services’ for more details on how these thresholds were chosen. Fish size has no informative meaning. c Graphical representation of the ecosystem services impaired by submerged plant growth, as reported in a

Full size image

Thresholds for boat traffic

In many freshwater ecosystems a large portion of boat traffic will consist of recreational vehicles. Aquatic macrophytes will become a nuisance when they entangle the propellers of motorboats or hit and wrap around the keel or swards of sailing boats. This results in reduced recreational use of the location (Eiswerth et al. 2000; Dodds et al. 2009). Threshold levels thus depend on the draught of the boats present at a location. For our case study, a cover of as low as 5% within the navigational area is suggested to cause nuisance for boating if macrophytes grow tall enough to come in contact with the boat (HISWA, personal communication). A survey of nearly the entire Dutch recreational fleet (for details, see WaterrecreatieAdvies 2005) and advice provided by HISWA made it possible to roughly group the majority of vessels into two groups and estimate the draught of the most common recreational boats. The report by WaterrecreatieAdvies (2005) described the number of vessels, type and estimated size of boats present in the Dutch recreational fleet. The first group consists mainly of small boats (up to 1–1.25-m draughts) stored on land, but still frequently used. The second group consists of vessels that are mainly docked at marinas or water sports associations and mainly consist of larger boats with larger draughts up until around 2 m. Because almost all boats will draw at least 0.5 m of water, all recreational boat traffic will have problems if macrophytes are present between 0 and 0.5 m below the water surface and cover at least 5%. The majority of the Dutch fleet will encounter problems when macrophyte canopy depth lies between 0 and 1 m deep. A smaller portion of the fleet in our case study (roughly half to two-third) still encounters problems when the macrophyte canopy depth is between 1 and 2 m below the water surface. When macrophytes deeper than 2 m are present only very few vessels will encounter problems, even when the macrophyte cover is high.

Thresholds for fishing

Lines getting entangled in vegetation pose a major problem of submerged macrophytes for fishermen. Catch is lost and/or lines can break, leading to loss of gear and pollution of the environment. High vegetation density results in reduced recreational use of affected sites (Eiswerth et al. 2000). According to The Netherland’s Sport Fishing Association (SVN) the recreational fishermen regard a macrophyte cover of between 10 and 40% as optimal for fishing in our case study (Peters and Van Emmerik 2013). In the USA, a similar cover of between 20 and 40% has been suggested as optimal for a stable fish population (AERF 2005). A cover of 50% and higher is considered a nuisance (SVN personal communication). Depending on the species of fish targeted, different fishing methods will be applied. When fishing for bottom-dwelling species macrophytes of only a few centimetres tall can already cause problems. When fishing for species that prefer the open water macrophytes can grow much taller before becoming a nuisance. As a waterbody or waterway will most likely be used to fish for both types of fish, a macrophyte cover of 50% or higher is classified as a nuisance for fishing independent of canopy depth in our case study. Macrophytes can also hamper fishing by causing problems for boat navigation, but these problems are described under the section concerning ‘thresholds for boat traffic’.

Thresholds for swimming

When macrophytes reach very high densities they can become dangerous for swimmers, the shoots can entangle arms and legs leading to dangerous situations. However, even when macrophyte densities are not dangerously high, many people are still deterred from swimming by submerged macrophytes touching their skin. At what cover and canopy depth people will find submerged macrophytes a nuisance will depend on personality and culture. Recreational swimmers in The Netherlands generally do not mind the presence of ‘some vegetation in the water’ (Goossen et al. 1997). Taking the average length of humans into account, plants present at 2 m or deeper will almost never be considered a nuisance in our case study. Plants present at 1.5–0 m depth might be considered a nuisance as people are likely to be able to touch them while swimming. As the presence of some vegetation will still be considered tolerable and not deter people to swim in waterbodies or waterways on a warm day, the cover at which macrophytes will be considered a nuisance has been arbitrarily set at 10% in our case study, but should be adjusted to reflect local views.

Thresholds for hydrodynamics

One of the main functions of waterways in particular, but also of hydropower reservoirs, is the transport of water (either to drain or to supply). Everything present in the water column will create resistance to the water flow and thereby reduce discharge capacity of the system. Massive development of submerged macrophytes can thereby severely reduce the amount of water that can flow through the system in a given time, increasing the chance of flooding (Bal et al. 2011). Several authors examined the hydraulic resistance of submerged macrophytes and the effects on water flow; however no clear threshold level for cover and canopy depth above which the vegetation will cause a significant problem was reported. The nuisance threshold will depend on the characteristics of the macrophyte species (e.g. flexibility), minimum flow capacity requirement of the water system at a specific time, pattern of the vegetation in the aquatic system and the bottom morphology of the waterway (Green 2005; Vereecken et al. 2006; Bal et al. 2011). Modelling indicated increased flow resistance at vegetation blockage of a channel of more than 50% (for details and model conditions, see: Green 2005). Empirically, more than 42% PVI (i.e. water volume inhabited by plants) did increase flow resistance under experimental conditions (Vereecken et al. 2006). However, up to 60% of vegetation in an artificial channel showed a flow resistance (Vereecken et al. 2006) that would not necessarily lead to flooding problems in rivers (Bal et al. 2011). Summarising, a PVI of 50% appears to be a good starting point threshold level at above which macrophytes will cause hydrological problems in our case study. It should be noted that the spatial pattern of the vegetation influences the resistance it will inflict on water flow (Green 2005; Vereecken et al. 2006; Bal et al. 2011). Furthermore, plant fragments can also easily clog pipes and pumps used for cooling water or irrigation (Richardson 2008). However, no general threshold value can be given as this will depend on the local specification of the pump (e.g. size and power), diameter of the pipes, mesh size of debris screens, etc.

Compiling the framework

Different ecosystems may provide different ecosystem services, the first step in classifying nuisance should thus be to identify which services are provided by the ecosystem at hand. Here we will show how to develop a classification scheme for a model shallow lake in The Netherlands that provides all the ecosystem services discussed above. We combine all these threshold values for cover and canopy depth from the separate ecosystem services and create one classification scheme for nuisance submerged macrophytes for our case study. We chose an ecosystem with a water depth of 2 m as an illustrative case study, but the same method can be applied in waterbodies and waterways with different depths. Threshold nuisance submerged macrophyte levels of cover and canopy depth for boat traffic, fishing and swimming in the case study can be used in the classification scheme without having to be transformed using our site-specific water depth (B, F, S in Fig. 1a, respectively). Because our classification framework requires both canopy depth and cover as main factors, the threshold PVI value of the hydrodynamics category needs to be converted to these two parameters. The PVI can be calculated by dividing submerged macrophyte volume by water volume. Submerged macrophyte volume can be calculated by multiplying plant height with plant cover. To calculate the macrophyte canopy depth and cover at the threshold level PVI for nuisance we thus need to know the water depth. A threshold PVI of more than 50% is reached at any combination of 75–100% plant cover and canopy depth of 0.5 m below the water surface or less in our schematic (H in Fig. 1a). Even though the PVI is easy to calculate in theory, determining the PVI of plants under field conditions can be difficult. Variation in shoot densities within a plant patch can make it difficult to accurately assess true cover. Furthermore, water flow can reduce plant height by pushing the shoots downwards making actual height estimates challenging. Overall, submerged plants will not cause problems in our case study when they lie on the bottom and cover less than 50% of the surface or when they cover less than 5% of the area, irrespective of canopy depth at that site (Fig. 1a).

Discussion

Nuisance problems, vegetation characteristics and nuisance species

A wide variety of problems for human use and ecological functioning of ecosystems caused by submerged macrophytes has been reported. Four main categories could be identified for human-related problems: problems for boat traffic, fishing, swimming and hydrodynamics. Each category (i.e. ecosystem service provided) has a unique threshold level at which plants become a nuisance. Threshold levels for nuisance of measurable vegetation parameters were largely absent in the scientific literature, but vegetation cover and canopy depth below the surface were most often used to describe nuisance. In our classification framework, we therefore combined plant cover with plant canopy depth as the main factors determining if a macrophyte stand will cause nuisance for a specific ecosystem service, or not.

Dozens of submerged macrophyte species were recorded in the scientific literature as causing problems for humans. Most of these species are known to possess a vertical growth strategy and some are notorious for forming surface canopies (e.g. E. densa, Elodea spp., H. verticillata, L. major, M. spicatum, Potamogeton pectinatus and P. crispus). Species with these characteristics likely pose the biggest risk. Many of these species also require this vertical growth to reach the water surface in order to flower (see Duarte and Kalff 1990). However, several other species reported to cause nuisance are not readily associated with this type of growth (e.g. Charophytes and Najas flexilis, Wingfield et al. 2006; Table 2). The list of species that can cause nuisance presented here is not exhaustive. Many more species have the potential to become a nuisance when environmental conditions allow for excessive growth. Environmental conditions enabling nuisance growth will most likely vary between species (reviews by Lacoul and Freedman 2006; Bornette and Puijalon 2011).

Applying and improving the nuisance classification framework

Integrating ecosystem morphology

We used an ecosystem with a depth of 2 m as an illustrative case study to convert nuisance threshold PVI values into canopy depth and cover. Because of the shallow depth of these ecosystems, light can often reach the bottom enabling submerged macrophyte growth throughout the ecosystem (e.g. Trebitz et al. 1993; Hilt et al. 2006; Pot and Ter Heerdt 2014). Therefore, these shallow ecosystems may be particularly vulnerable for problems concerning nuisance growth of macrophytes (e.g. Burlakova and Karatayev 2007; Alwin et al. 2010). The classification scheme can easily be adapted for waterbodies and waterways with different depths, as the nuisance threshold value itself remains unchanged. However, the deeper a system becomes, the less likely macrophytes are to reach the threshold value for canopy depth, and thus nuisance, because of the physiological limitations of the plants (Lacoul and Freedman 2006).

Integrating macrophyte species characteristics

Many submerged macrophyte species have been a nuisance, but several species have been reported more often than others. This could be due to specific species’ traits such as the ability to quickly grow tall and form surface canopies or show strong lateral expansion, thereby increasing the likelihood of becoming a nuisance, although species without the same traits can also become a nuisance, but perhaps not as often. Including information on specific plant species on potentially relevant plant traits such as toughness (Bociag et al. 2009), flexibility or maximum shoot density could thus further improve the classification framework by including species-specific threshold levels at which macrophytes are considered a nuisance. The impact species-specific traits can have on the nuisance threshold levels can, for example, be included in the framework by adding a species specific correction factor (i.e. ranging from 0 to 1) to the general threshold levels. Here, a very mechanically strong species will have a factor of 1 and very weak species will have a lower correction factor.

The present study focussed on submerged freshwater macrophytes, but similar problems concerning nuisance growth of species with different growth forms, such as floating species, have been reported (e.g. Spencer et al. 2006; Arthur et al. 2007; Wu and Wu 2007; O’Sullivan et al. 2010; Perna et al. 2012). In contrast to submerged species, floating leaved species such as Eichhornia crassipes or Trapa natans, but also emergent species such as Ludwigia spp. or Phragmites australis, may be even more likely to cause nuisance problems, because they inherently grow on or through the water surface and thus have a high chance of causing nuisance problems to ecosystem users. Due to the general nature of the nuisance problems and the vegetation characteristics used to classify nuisance, our proposed classification can easily be adapted and extended to include macrophyte species with different growth forms. For example, the plant canopy depth factor will not be useful in the classification framework when applied to floating leaved species (i.e. canopy depth is always 0). However, the cover threshold values of the framework are still valid.

Integrating varying ecosystem services and local perception

While performing the review, it became apparent that it was impossible to create one threshold value of cover and depth above which aquatic plants will be perceived as a nuisance (e.g. Suren 2009) and our classification will thus benefit from including site-specific information on perception of nuisance. For example, Chambers et al. (1999) summarised that the tolerable upper limit of aquatic plant cover, before complaints are filed and management is triggered, lies between less than 1% and up to 50% in Canada. This is in agreement with our case study and shows the large range this threshold can have because of local views and usage requirements of the different assigned ecosystem services. We therefore propose that managers and scientists tailor the framework to the ecosystem at hand by (1) identifying the ecosystem services provided by the ecosystem, (2) determining at what per cent cover and canopy depth the plants will impair each of the services provided and (3) assessing whether a corresponding reduction in macrophytes is beneficial for the ecosystem or whether it will lead to conflicts among different services with different plant requirements. A large reduction in submerged macrophyte PVI can potentially have detrimental effects on the ecosystem and the ecosystem services provided.

Aquatic plant problems versus services

Eliminating nuisance submerged vegetation via management may reduce anthropogenic problems for usage of aquatic systems, but many ecosystem services also rely on the ecosystem functions performed by submerged plants, such as maintaining good water quality and healthy fish populations.

Water quality is very important for recreation (Goossen et al. 1997). Fishing, hunting and bird watching also benefit from submerged plant presence, as these plants may attract waterfowl and increase fish stocks (Noordhuis et al. 2002; AERF 2005; O’Hare et al. 2007; Hansson et al. 2010; Peters and Van Emmerik 2013). Removing too many plants can increase the risk of complete loss of submerged vegetation, development of potentially toxic phytoplankton blooms and thus indirectly jeopardise several human uses of the system (Van Nes et al. 2002; Dodds et al. 2009; Kuiper et al. 2017). However, it is unknown how much plant volume is actually needed for a stable clear water ecosystem (Hilt et al. 2006), because the minimum PVI required will differ per ecosystem and depends on factors including: macrophyte traits; nutrient status of the system; presence, type and abundance of fish; and many more (e.g. Søndergaard and Moss 1997; Hilt and Gross 2008; Bakker et al. 2010). Most likely, PVI in temperate freshwater ecosystems should be higher than 15–30% to maintain a clear and macrophyte-dominated lake (Søndergaard and Moss 1997; Hilt et al. 2006; Nakamura et al. 2008; Sondergaard et al. 2010), but no threshold PVI level was found for waterways. However, a too high PVI of (close to) 100% can be detrimental for water quality and ecosystem stability, especially when extensive surface canopies are formed. This can result in very low oxygen concentrations near the bottom in the macrophyte stands (Schwarz and Howard-Williams 1993; Miranda and Hodges 2000; Nakamura et al. 2008).

Next to increasing water clarity, submerged vegetation may additionally have a direct and positive impact on one ecosystem service in particular: fishing, likely because the aquatic plants provide food and habitat for many animals, including fish (Carpenter and Lodge 1986), thereby maintaining a stable fish population. Similar to the provision of clear water, no single optimal value of macrophyte PVI for the fish populations was found in the scientific literature. Macrophyte cover values in the region of 20–60% have been suggested to be beneficial for fish populations and fishing (Dibble et al. 1996; AERF 2005; Peeters and Van Emmerik 2013). Very high macrophyte densities can result in low oxygen levels and potentially decrease fish populations (Miranda and Hodges 2000). However, this optimal macrophyte cover may not be true for all fish species or aquatic ecosystems (e.g. stratified temperate lakes: Cheruvelil et al. 2005).

Using these suggested PVI and cover values and using our case study ecosystem with a depth of 2 m as an example, we developed an additional classification scheme showing when too few or too many plants are present in the lake to maintain the two beneficial ecosystem functions discussed (Fig. 1b). Because these threshold values may vary among different ecosystems (Murphy 1988), we advise performing an ecosystem analysis to tailor the values to specific ecosystems.

Spatial heterogeneity

It now becomes apparent that it is nearly impossible to use an entire shallow lake for the anthropogenic functions discussed while keeping clear water in the ecosystem in the plant-dominated state (see Fig. 1a versus b). This seems only possible if abiotic factors prevent nuisance submerged plant growth all together (e.g. very low nutrient concentrations, high flow velocity or rocky substrates). However, this does not mean that ecosystem managers will have to choose for either (1) an aquatic system in the plant dominated state, but without the possibility of unhampered anthropogenic functions, or (2) a system without any submerged macrophytes and a high chance of poor water quality. It merely stresses the importance of incorporating whole ecosystem functioning and spatial aspects into the management plans for selected human uses of the system (Finlay and Vogt 2016). This is especially true because it is not always possible to assign the different ecosystem functions and services to different aquatic ecosystems.

So far, we have not addressed the spatial heterogeneity of aquatic plants and ecosystem services, but many ecosystem services do not require the whole aquatic ecosystem. Swimming, for example, will mainly be restricted to the area surrounding a beach or an easy access point and will remain close to the shoreline. So, the canopy depth and cover of the vegetation needed to prevent nuisance to the swimmers will only apply locally. Similar arguments can be made for boating and fishing. Consequently, management only needs to take place in areas where macrophyte growth actually causes problems (Finlay and Vogt 2016). Other parts of the ecosystem can still be occupied by submerged macrophytes to maintain their functions for the ecosystem as a whole. Furthermore, submerged macrophytes themselves often show a patchy distribution in aquatic ecosystems and activities could thus be allocated towards the sites with fewer plants. If the underlying causes of nuisance submerged macrophyte growth cannot be removed, harvesting nuisance aquatic plants may be a suitable management method to temporarily alleviate local problems (Quilliam et al. 2015; Finlay and Vogt 2016; Hussner et al. 2017). An additional advantage of this method is that it removes nutrients that are incorporated in the plant biomass from the ecosystem, which can be used for a wide variety of useful applications, for example as agricultural fertiliser or food for cattle (Edwards 1980; Quilliam et al. 2015).

Conclusions

Our classification framework shows that the number of plants considered to be a nuisance depends on the services the ecosystem fulfils. The framework combines vegetation cover and canopy depth as the main factors for the quantitative classification of nuisance versus non-nuisance submerged vegetation. The classification framework can be used to define and evaluate lake and waterway management goals, as it is possible to quantitatively define nuisance and desirable vegetation dimensions, depending on the ecosystem service provided. The use of our classification framework in future research also enables researchers to compare individual studies where nuisance macrophyte growth is reported as well as review the underlying causes of nuisance growth.

Whereas submerged macrophytes can become nuisance vegetation for multiple ecosystem services, we stress that submerged macrophytes simultaneously provide essential services in shallow freshwater ecosystem, for example, maintaining stable clear water conditions. Therefore, we suggest that lake management should strive for spatial differentiation of human activities and plant growth if the underlying mechanisms enabling nuisance cannot be removed.