Novel methods for analysing benthic structure
(mussels and macrophytes) of freshwater systems
Source: www.hef.ru.nl
O. (Olivier) van Altena 11908521 Examinator: dr. Arie Vonk
Assessor: dr. Harm van der Geest (IBED) Daily supervisor: Elmar Becker Msc (IIS)
In cooperation with: University of Amsterdam IBED
1
Content
Abstract ... 2
Introduction ... 3
Background: The role of macrophytes and filter feeders in reducing turbidity in freshwater systems ... 3
Method ... 5
Study area ... 5
Location ... 5
Physical and chemical parameters ... 5
Biological components ... 5
Techniques ... 5
Side scanning sonar ... 5
Multibeam Echosounder ... 6
Quantitative video analysis ... 6
Data analysis ... 6 Results ... 7 Descriptive stats ... 7 Sonar data ... 7 Dreissenid results... 9 Macrophytes results ... 10 Discussion... 11
Dreissenid mussels and Side scanning sonar ... 11
Dreissenid mussels and depth ... 11
Macrophytes and Side scanning sonar ... 11
Macrophytes and depth ... 11
Outcomes ... 12
Improvements ... 12
Recommendations ... 12
Conclusion ... 13
References... 14
2
Abstract
Dreissenid mussels and macrophytes both play an important role in freshwater lake ecosystems. One example is their role in the regulation of water quality and transparency. Both organisms are also an important part of the food chain by consuming nutrients and being consumed by other organisms. However, dreissenids and macrophytes can also form a nuisance for humans in certain conditions. Because of their great importance and their potential production of hinder, it is important to monitor these organisms. A novel way of doing this is recreational grade sonar. However, before being implemented, the validity should be investigated further. Therefore, this research investigates the aims to determine whether sonar can be used to accurately determine the dispersion and densities of dreissenid mussels and macrophytes, and be used as a tool for ecological monitoring? This was done by comparing roughness scores and depths, measured by sonar, with counted dreissenid mussel and macrophyte presence and densities. The results show that there is a significant positive correlation between the observed dreissenid densities and the roughness scores. A significant negative correlation between dreissenid density and depths, but no significant correlation between macrophyte presence and the roughness scores or depth. Therefore, it can be concluded that sonar can be used as a tool for monitoring dreissenid mussel densities, but further research is needed for sonar to be used as a tool to monitor macrophyte densities.
3
Introduction
In freshwater lake ecosystems, benthic organisms such as dreissenid mussels and macrophytes play an important role in different sections of the ecosystem (Jeppesen, Sondergaard, Sondergaard & Christofferson, 2012; IJff, 2014). Both macrophytes and dreissenid mussels are vital in the regulation of water quality and transparency of the water (Kuiper et al., 2017).
Background: The role of macrophytes and filter feeders in reducing turbidity in freshwater systems Dreissenid mussels and macrophytes can promote an increase in water transparency, in multiple different ways. Macrophytes can mediate the increase in transparency of the water (Genkai-Kato, 2007). Submerged macrophytes can act as a refuge for zooplankton (Jeppesen, Jensen, Søndergaard & Lauridsen, 1999). This will cause a decrease in the predation of planktivorous fish on zooplankton. Therefore, an increase in grazing of zooplankton on phytoplankton can occur (Genkai-Kato, 2007). In general, phytoplankton tend to increase the turbidity of the lake water (Radke & Gaupisch, 2005). This means that an increase in the grazing of zooplankton on phytoplankton will increas the transparency of the water. However, planktivorous fish species can also use the submerged
macrophytes as a refuge against the predation from the piscivorous fish species (Genkai-Kato, 2007). This implicates that the effect of increasing transparency induced by the macrophytes will be greater with the appearance of piscivorous fish to control the planktivorous fish (Genkai-Kato, 2007). Macrophytes are also able to affect phytoplankton concentrations by decreasing the phosphorus recycling from sediments (Genkai-Kato & Carpenter, 2005). Phosphorus recycling occurs when phosphor binds to iron under oxic conditions and is released into the water when the layer of water above the sediment is deoxygenated (Genkai-Kato & Carpenter, 2005). Once released in the water, the phosphor can be consumed by phytoplankton. Macrophytes can decrease this recycling rate by consuming the phosphorus themselves when the phosphorus is still trapped inside the sediment (Genkai-Kato & Carpenter, 2005).
Besides increasing the transparency of the water, macrophytes play an important role in the trophic interactions between prey and predators. Submerged macrophytes act as habitat and shelter for many prey (Diehl & Kornijów, 1998). Because of this reason, macrophytes often serve as nursery grounds for diverse fish species (Gasith & Gafny, 1998).
Besides macrophytes, dreissenid mussels also have a positive effect on the increase of the water transparency (Stroom, 2016). Dreissenids are filter feeders and therefore can improve water transparency by filtering phytoplankton and other small particles out of the water
(Hoogheemraadschap rijnland, 2015). The filter capacity of dreissenid mussels is dependent on the flow velocity, the nutrient concentrations and nutrient composition (IJff, 2014). Due to the
consumption of phytoplankton, quagga mussels are competing with zooplankton for this resource (IJff, 2014). Because of the increased water transparency, caused by the presence of dreissenid mussels, macrophytes are able to increase their densities and dispersion (IJff, 2014). This increase occurs especially when nutrient concentrations in the sediment are high. Dreissenid mussels are also capable of altering the nutrient concentrations by consuming nutrients from the water in the form of particulates and by excreting faeces containing nutrients keeping it near the benthos (IJff, 2014). Like macrophytes, dreissenid mussels also play a role in the trophic system of freshwater lake ecosystems. As previously mentioned, dreissenids consume phytoplankton and therefore compete with zooplankton. There is also some predation on dreissenid mussels in the Netherlands, mainly performed by certain some fish and duck species (IJff, 2014).
4 Despite the important function dreissenid mussels and submerged macrophytes have in freshwater lake ecosystems, they can also constitute a nuisance. For instance, juvenile dreissenid mussels attach themselves on hard surfaces like rocks and pipes, causing obstacles for human use when densities grow too large (Hoogheemraadschap Rijnland, 2015). Macrophytes can form a possible obstruction for recreational and commercial boats and people can experience hinder or irritations while swimming through them (Kuiper et al., 2017). For these reasons, macrophytes are sometimes removed from freshwater bodies (Kuiper et al., 2017).
Because dreissenid mussels and macrophytes play such an important role in the regulation of the transparency and water quality in freshwater lake ecosystems , it is important to measure the densities and distribution of these benthic organisms and display them on maps. Especially when these organisms threaten to become a problematic factor in the relationship between humans and the ecosystem. Conservation strategies and management can be performed more effectively with the use of these measurements. The traditional methods for these measurements and samplings, like scuba diving, corers and dredges, are not the most efficient and cheap methods to use for measuring the distribution and densities of benthic organisms (Sánchez-Carnero, Rodríguez-Pérez, Zaragozá, Espinosa & Freire, 2014).
More efficient methods are acoustic methods that scan large areas of the lake floor
(Sánchez-Carnero et al., 2014). Two examples of those acoustic methods are the Side Scan Sonar (SSS) and the Multibeam Echosounder (MBES). One of the areas where Side Scan Sonar and Multibeam
Echosounder have been used are the lakes of the Kagerplassen in the Netherlands. SSS is mainly used to measure the roughness of the lake bottom surface. Meanwhile, MBES is used to measure the depth of the lakes. Potentially, these roughness scores can be used to determine the presence of macrophytes and densities of dreissenid mussels.
However, it is not quite known how precise these estimates are. Therefore, the aim of this research is to investigate whether sonar is a valid method to estimate the presence and densities of benthic dreissenid mussels and macrophyte patches. This leads to the research question: Can sonar be used, by displaying the dispersion and densities of dreissenid mussels and macrophytes, as a tool for ecological monitoring? It is hypothesized that with sonar techniques the dispersion and densities of dreissenid mussels and macrophytes can be determined and that it is therefore a useful tool that can contribute to the monitoring and management of such benthic organisms. To answer this research question, the following sub questions will be answered.
• Do Side scanning sonar values correspond with the density of dreissenid mussels measured by qualitative video analysis?
• How does the Side scanning sonar correspond with the presence of macrophytes? • How does depth influence the density of dreissenid mussels and the presence of
5
Method
Study area Location
The study area used in this research will be the lakes of the Kagerplassen. They are located just above the city of Leiden and cover an area of 235ha (Van Puijenbroek, Janse & Knoop, 2004). The lakes of the Kagerplassen consist of eight different smaller waterbodies and are connected with each other through canals (Factsheet Kagerplassen). The first pools arose from the lowest points due to a sudden sea level rise around the year 600 (Factsheet Kagerplassen). The remaining part of the lakes were used for sand extraction, which has transformed the lakes into their current state (Factsheet ‘t Joppe). Currently, the lakes of the Kagerplassen serve as a recreational and water storage area, and are used to regulate the waterflow (Factsheet Kagerplassen).
Physical and chemical parameters
The Joppe is the largest waterbody of the Kagerplassen and is relatively deep in comparison with the other parts. There is a high nutritional input from the surrounding area. This is because the area around the lakes of the Kagerplassen is highly populated and linked to the presence of agricultural land around the area (Van Puijenbroek, Janse & Knoop, 2004). Wastewater treatment plants are a large contributor to the nutritional emission load into the lake areas (Van Puijenbroek, Janse & Knoop, 2004).
Biological components
The main species of dreissenid present in the lakes of the Kagerplassen is the Dreissena bugensis or the quagga mussel (Hoogheemraadschap rijnland, 2015). This species of dreissenid originates from the Ponto-Caspian region, and have accidently been introduced in the Dutch waters, where they have been expanding their range since then (Hoogheemraadschap rijnland, 2015). Because Dreissena
bugensis is a more efficient disperser, they suppress the dispersion of the already present Dreissena polymorpha (Noordhuis, 2014).
In the lakes of the Kagerplassen, there are a couple of different species of submerged macrophytes present. One of the more abundant and relevant species are Elodea nuttallii and Nuphar lutea. E.
nuttallii is resistant against low light intensities, but is also able to grow under light intensities of
1100 mmol m-2 s-1 . This allows E. nuttallii to grow in a large variety of lake depths(Erhard & Gross, 2006). Nuphar lutea occur in slow flowing, shallow freshwater bodies and has floating leaves (Ivanova, Philippov, Kulichevskaya & Dedysh, 2018).
Techniques
Side scanning sonar
The density of the quagga mussel in the lakes of the Kagerplassen, side scanning sonar is used to measure the roughness of the bottom of the lake. With these roughness values, local densities of dreissenid mussels can be estimated. The Hoogheemraadschap of Rijnland provides this research with the necessary side scanning sonar data. The sonar that makes measurements for the data used in this research, uses a 600 kHz Edgetech 4125 SSS system. During scanning, the sonar is held 0,7 m under the water surface and had a range of 40 m. Side scanning sonar works by sending an acoustic pulse and receiving the echo. By linking the intensity of the echo with the location, a representation can be formed with unique intensity values for each coordinate. This representation was made into a map using the program ArcMap. In ArcMap, a layer with the intensity scores and a layer with the depth values was added.
6 Multibeam Echosounder
To measure the depth of the lake at certain points, another sonar system is used. The Multibeam Echosounder. This sonar uses the interferometric multibeam system GeoSwath Plus. The sonar consists of two transducers of 250 kHz on a v-shaped frame. To compensate for the movement of the boat and the waves, an Octans 100 movement sensor is installed in the frame. By scanning strokes of the lake, a surface covering image can be created. Just like the SSS data, the MBES data is converted into a map with ArcMap.
Quantitative video analysis
For the quantitative video analyses, a Go-Pro waterproof action camera is attached to a pole and lowered into the water. From each video of a location, one image is chosen to function as sample for the analysis. This image is chosen for its representativeness of the area visible on the record.
Subsequently a screenshot is taken from this image and posted in word. The total number of dreissenid mussels in the image are counted. The density is then calculated in the total number of dreissenid mussels per squared meters.
The numbers of mussels in the screenshot is converted to the density of mussels with the use of a grid which is attached on the Go Pro stick. The size of the grid is known, which is 15 by 15 cm, and is visible on the Go Pro records. By using the size of the grid, the size of the screenshot can be
calculated and therefore the surface of the screenshot. The density of the dreissenid mussels is then calculated with the following formula (in mussels/m2): #mussels/((length screenshot(15/length grid))(width screenshot(15/length grid))). These density values are subsequently used in the statistical analysis and compared with sonar scores from the side scanning sonar and the depth values. This comparison is done with the map made in ArcMap. The coordinates from the Go Pro records were added in the program and became therefore visible on the intensity and depth layers. This creates the possibility to link the right intensity scores and depth values with the right dreissenid densities.
The screenshots are also used to determine the presence and distribution of macrophytes. Every screenshot where macrophytes are present, gets the value 1. And every screenshot where no
macrophytes are present at all, gets the value of 0. Just like the dreissenid data, the macrophyte data is compared with the same sonar scores and depth values retrieved from the maps made in ArcMap. Data analysis
For the data analysis, a correlation test is executed in the program R. To perform such a test, its corresponding assumptions have to be confirmed. To check whether the response, in this case the dreissenid density, is normally distributed, a shapiro wilk test is used. The outcome of this test shows that the dataset is not normally distributed. This means that the Spearman’s rank correlation test must be used instead of the common correlation test. These steps are performed on the data combinations of dreissenid densities and sonar scores and on dreissenid densities and depth. To test whether the sonar scores can predict if macrophytes are present or not, a generalised linear model test is performed. This is done with the glm() function. The feature, family = binomial is added on this function. Subsequently, the summary() function is performed on the previous glm() function. This shows the formula of the regression and the corresponding p-value. The same procedure for the comparison of macrophytes and sonar is also done for comparison of macrophytes and depth. All analyses were done with R (3.6.1) and R studio (1.1.419)
7
Results
Descriptive stats
In total, the number of analysed data points is fifty-seven. In thirty-four of those data points, there was a presence of dreissenid mussels and the mean dreissenid mussels per square meter of all these data points is 529,7 and the standard deviation is 733,5. In eleven of the total screenshots,
macrophytes were present. There was a great variance in the number of individual macrophytes visible on the screenshots. Some screenshots contained just one individual and others were almost entirely filled with macrophytes.
Sonar data
The scores of the SSS data and the MBES sonar data are shown in figures 1 and 2. All the used data points in this research are visible in these maps and labelled with its corresponding ID number. The sonar in these maps are visible as a flat covering layer. Figure 1 shows the SSS data. With from blue to red increasing values up to sixty-three. Most of the high values are located in channels used by boats and around the deeper parts of the lakes, with the Joppe as an example . Figure 2 shows the MBES data. With from yellow to green increasing values. The deepest part in the area is the Joppe. Other slightly deeper parts are the channels used by boats and one area in the middle of the centre lake.
Figure 1, Map of the lakes of the Kagerplassen with a flat covering layer containing the sonar data and the ID labelled points where the records have been recorded. The values increase from blue to red.
8
Figure 2, Map of the lakes of the Kagerplassen with a flat covering layer containing the depth data and the ID labelled points where the records have been recorded. The values increase from yellow to blue.
9 Dreissenid results
The Shapiro Wilk test showed the following results, W = 0,75337 and value = 2,007e-08. The p-value is not larger than 0,05 and the response is therefore not normally distributed. Figure 3 shows the distribution of the dreissenid density data.
Figure 3, The distribution of the dreissenid density data.
The relation between dreissenid density and the sonars scores is shown in figure 4, with the fitted regression line of the correlation between the dreissenid density in the lakes of the Kagerplassen and the sonar scores made by the SSS. The Spearman’s rank correlation test gave a correlation of
0,4989581 and a p-value of 3,898e-05. This means that the correlation given by the test is significant, because the p-value is lower than 0,05.
Figure 4, The regression line from the correlation between the dreissenid densities and the sonar scores. Density = 102,36 + 29,92*sonar
10 In figure 5, you can see the regression of the correlation between dreissenid densities and the
measured depth. The Spearman’s rank correlation test gave a correlation of -0.3490646 and a p-value of 0.007785. This also means that the correlation is significant, because the p-p-value is below 0,05.
Figure 5, The regression line from the correlation between the dreissenid densities and the depth of the lakes.
Macrophytes results
The generalized linear model analysis of the macrophyte data and the sonar score resulted in the following regression, macrophyte presence = -1,86523 + 0,02900 *sonar score. The analysis also gave the p-value of 0,41023. This value is not below 0,05. This means that this regression is not significant. The results from the generalized linear model analysis of the macrophyte data and the depth showed the regression, macrophyte presence = -2,0452 + 0,2461*depth. With this regression, the analysis gave a p-value of 0,763. This p-value is also is not below 0,05, that is why this regression is, just like the previous one, not significant.
11
Discussion
Dreissenid mussels and Side scanning sonar
So from what we have seen, dreissenid densities have a positive correlation with the sonar scores. This is in line with the expectations for this part of the results. It is commonly known and proven in previous research that mussels increase the substratum roughness (Quinn & Ackerman, 2014). This roughness is measured by the sonar and therefore you would expect a higher sonar score with a higher mussel density.
Dreissenid mussels and depth
The results showed a negative correlation between dreissenid densities and depth. This is not what was expected, because previous research has shown that quagga mussels, who are the main dreissenids in the lakes of the Kagerplassen, can occur at depths of 100m (Higgins & Zanden, 2010). Previous research has shown that the depth where dreissenid mussels occur is determined by a couple of ecological factors, like temperature or the substratum (Farr & Payne, 2010). For dreissenids to settle, a hard surface substratum is necessary. The depth of a fitting substratum is in this case determent for the depth where the mussels occur (Verhofstad, Grutters, Velde & Leuven, 2013). Thereby, when two different competing species of dreissenids occur together in the same waterbody, both species will eventually zonate in depth (Verhofstad, Grutters, Velde & Leuven, 2013). This being said, dreissenids tend to search for the most optimum circumstances and therefore also the depth where those circumstances are present (Verhofstad, Grutters, Velde & Leuven, 2013). This phenomenon could be one of the reasons the correlation between the dreissenid densities and depth are negative. By monitoring the ecological factors that determine the most optimal depth, more precise estimates of the dispersion of dreissenid mussels can be made in the future. Macrophytes and Side scanning sonar
It was expected that the relation between macrophytes and sonar scores would be positive. This is based on the fact that macrophytes tend to decrease the flow velocity of the water body they are present in (Madsen, Chambers, James, Koch & Westlake, 2001). This increases the opportunity for sediments to settle and therefore increase the roughness of the substratum (Madsen et., 2001). However, the logistic regression between macrophytes and sonar scores was not significant. One of the explanations for this result is that there is a great shortage of macrophytes in the Kagerplassen. This can also be seen in the results, eleven out of fifty-seven screenshots contained macrophytes. It may have been the case that there simply are too little macrophytes present. Therefore, the effect on the flow velocity may not have been significant because of the fact that the densities of the macrophytes are too small.
Macrophytes and depth
Just like the macrophyte and sonar results, the logistic regression between macrophytes and depth is not significant. This can also be explained by the fact that there are not that many macrophytes present. In theory, a much larger area in the Kagerplassen would be suitable for macrophytes to grow when you look at the depth. The lakes have almost everywhere the same depth, see figure 2. There are some patches with macrophytes at the most common depth and if you only consider depth as a factor, all the places with this depth should be suitable for macrophyte growth (Søndergaard et al., 2013). This is not the case and this means that there are other limiting factors. Investigating those factors can lead to a possible causation for the low macrophyte densities.
12 Outcomes
The outcomes from these results is that for now, dreissenid densities do correspond with SSS values and can be estimated with the depth. However, when the depth is used to estimate the mussel density, other important ecological factors must be taken into account to determine the optimal depth. The presence of macrophytes could not be determined by the SSS or depth data. Perhaps, a new formed method and a similar research in another lake can produce significant results. From these outcomes, you could state that sonar can be used for conservational purposes. However, especially in the case of macrophytes, future research is needed to further confirm this claim. Improvements
As previously mentioned, this research was completely executed with previously collected data provided by the Hoogheemraadschap of Rijnland. This was caused by the corona epidemic. A great improvement would be the collection of more recent data. This will improve the reliability of the results. It can also be an improvement to create video records that are focused on both dreissenid mussels and macrophytes. The video records used in this research were mainly focused on mussels. This also may be one of the causes the macrophyte results were not significant. Thereby, the quality of the screenshots was not always that good and it was sometimes hard to tell if there were
macrophytes and mussels or not.
Another improvement for this research is to develop a more efficient and precise method to count mussels in the screenshots. In this research, it was done by hand and the it is very likely that some mussels have not been counted. This could have been caused because of mussels who are located between the submerged macrophytes. These mussels were not always visible on the screenshots, but they do affect the sonar score. Dead mussels have likely also affected the sonar score and these were not included in the counting. A different method to collect the macrophyte data would also be a good improvement. This method should, just like the dreissenid method, be able to calculate the density or the coverage percentage of the macrophytes instead of the presence of the macrophytes. Recommendations
Recommendations for future researches are to measure the density of macrophytes instead of the presence of them. This will lead to a more fitting relation between macrophytes and roughness or depth. It is also beneficial to execute a similar research at different lakes or pools in the Netherlands to compare results and extend data sets. Comparing researches from different seasons could maybe display the seasonal variations. Macrophytes densities can alter between winter and summer and the sonar response possibly as well, therefore it is also useful to consider. This will eventually lead to a more complete and precise sonar method.
13
Conclusion
To conclude, the results from the correlation analysis showed that the correlation between the dreissenid density and the sonar scores is positive and significant. This means that a higher sonar score stands equal to a higher dreissenid density. Thus the conclusion from the first sub question is that the Side scanning sonar scores correspond in a positive regression with the density of the dreissenid mussels. It could be concluded from this statement that SSS is an useful tool to estimate the dispersion and distribution of dreissenid mussels.
The logistic regression analysis between the presence of macrophytes and the SSS data resulted in a non-significant result. This implicates that the answer to the second sub question is that the SSS does not correspond with the presence of macrophytes.
Figure 5 did show that there is a significant negative correlation between depth and the densities of the dreissenid mussels. This implicates that the densities of the mussels will probably diminish the deeper you go. This answers partly the third sub question. The relation between depth and the presence of macrophytes was, however, not significant. This implies that depth does not correlate with the presence of macrophytes at a certain coordinate. The answer to the third sub question would then be, that the depth of a waterbody does correlate in a negative linear regression with the density of dreissenid mussels. And that the logistic correlation between the depth and the presence of macrophytes is not significant and therefore the depth does not influence the presence of macrophytes in waterbodies.
With the sub questions being answered, the main research question (Can sonar be used, by displaying the dispersion and densities of dreissenid mussels and macrophytes, as a tool for ecological conservation?) can be answered. SSS is able to estimate the dispersion and densities of dreissenid mussels by scanning the roughness of the lake bottom, but is not able to do the same for macrophytes. On the other hand, MBES is able to give correct depth values for certain coordinates and the results have shown that dreissenid densities can also be estimated with the use of the depth. So it can be said that MBES also could help for the mapping of dreissenid mussels. However, just like the SSS, depth does not correlate with the appearance of macrophytes. This implies that MBES is also not suitable to map the dispersion of macrophytes. So for conservational purposes, sonars can be suitable to map the dispersion and densities of dreissenid mussels in freshwater lakes like the lakes of the Kagerplassen.
14
References
Diehl, S., & Kornijów, R. (1998). Influence of submerged macrophytes on trophic interactions among fish and macroinvertebrates. In The structuring role of submerged macrophytes in lakes (pp. 24-46). Springer, New York, NY.
Erhard, D., & Gross, E. M. (2006). Allelopathic activity of Elodea canadensis and Elodea nuttallii against epiphytes and phytoplankton. Aquatic Botany, 85(3), 203-211.
Farr, M. D., & Payne, B. S. (2010). Environmental habitat conditions associated with freshwater dreissenids.
Gasith, A., & Gafny, S. (1998). Importance of physical structures in lakes: the case of Lake Kinneret and general implications. In The structuring role of submerged macrophytes in lakes (pp. 331-338). Springer, New York, NY.
Genkai‐Kato, M. (2007). Macrophyte refuges, prey behaviour and trophic interactions: consequences for lake water clarity. Ecology Letters, 10(2), 105-114.
Genkai-Kato, M., & Carpenter, S. R. (2005). Eutrophication due to phosphorus recycling in relation to lake morphometry, temperature, and macrophytes. Ecology, 86(1), 210-219.
Higgins, S. N., & Zanden, M. V. (2010). What a difference a species makes: a meta–analysis of dreissenid mussel impacts on freshwater ecosystems. Ecological monographs, 80(2), 179-196. Hoogheemraad Rijnland. (n.d.) Factsheet: NL13_01 't Joppe [Fact sheet].
https://www.waterkwaliteitsportaal.nl/General/DownloadFile?path=CustomReports/December2017 Publiek/Oppervlaktewater/factsheet_OW_13_Hoogheemraadschap_van_Rijnland_2017.pdf
Hoogheemraad Rijnland. (n.d.) Factsheet: NL13_06_2 Kagerplassen [Fact sheet].
https://www.waterkwaliteitsportaal.nl/General/DownloadFile?path=CustomReports/December2017 Publiek/Oppervlaktewater/factsheet_OW_13_Hoogheemraadschap_van_Rijnland_2017.pdf
Hoogheemraadschap Rijnland. (2015). Waterkwaliteitsverandering in relatie tot de aanwezigheid van Dreissena mosselen. Retrieved from
https://www.rijnland.net/over-rijnland/wat-doet- rijnland/schoon-en-gezond-water/downloads-schoon-en-gezond-water/rapport-waterkwaliteitsverandering-in-relatie-tot.pdf/view?searchterm=dreissena
Ivanova, A. A., Philippov, D. A., Kulichevskaya, I. S., & Dedysh, S. N. (2018). Distinct diversity patterns of Planctomycetes associated with the freshwater macrophyte Nuphar lutea (L.) Smith. Antonie van Leeuwenhoek, 111(6), 811-823.
Jeppesen, E., Jensen, J. P., Søndergaard, M., & Lauridsen, T. (1999). Trophic dynamics in turbid and clearwater lakes with special emphasis on the role of zooplankton for water clarity. In Shallow lakes’ 98 (pp. 217-231). Springer, Dordrecht.
Jeppesen, E., Sondergaard, M., Sondergaard, M., & Christofferson, K. (Eds.). (2012). The structuring role of submerged macrophytes in lakes (Vol. 131). Springer Science & Business Media.
Kuiper, J. J., Verhofstad, M. J., Louwers, E. L., Bakker, E. S., Brederveld, R. J., van Gerven, L. P., ... & Mooij, W. M. (2017). Mowing submerged macrophytes in shallow lakes with alternative stable states: battling the good guys?. Environmental management, 59(4), 619-634.
15 Madsen, J. D., Chambers, P. A., James, W. F., Koch, E. W., & Westlake, D. F. (2001). The interaction between water movement, sediment dynamics and submersed macrophytes. Hydrobiologia, 444(1-3), 71-84.
Radke, R. J., & Gaupisch, A. (2005). Effects of phytoplankton-induced turbidity on predation success of piscivorous Eurasian perch (Perca fluviatilis): possible implications for fish community structure in lakes. Naturwissenschaften, 92(2), 91-94.
Sánchez-Carnero, N., Rodríguez-Pérez, D., Zaragozá, N., Espinosa, V., & Freire, J. (2014). Relative infaunal bivalve density assessed from split beam echosounder angular information. Oceanologia, 56(3), 497-522.
Søndergaard, M., Phillips, G., Hellsten, S., Kolada, A., Ecke, F., Mäemets, H., ... & Oggioni, A. (2013). Maximum growing depth of submerged macrophytes in European lakes. Hydrobiologia, 704(1), 165-177.
Stroom, J. M. (2016). Quaggamossels en hun dominantie in de Sloterplas. Waternet, Amsterdam. Van Puijenbroek, P. J. T. M., Janse, J. H., & Knoop, J. M. (2004). Integrated modelling for nutrient loading and ecology of lakes in The Netherlands. Ecological Modelling, 174(1-2), 127-141. Verhofstad, M. J., Grutters, B. M., Velde, G., & Leuven, R. S. (2013). Effects of water depth on
survival, condition and stable isotope values of three invasive dreissenid species in a deep freshwater lake.
Quinn, N. P., & Ackerman, J. D. (2014). Effects of near‐bed turbulence on the suspension and settlement of freshwater dreissenid mussel larvae. Freshwater Biology, 59(3), 614-629. IJff, S. (2014). De quaggamossel in Nederland, een vloek of een zegen. STOWA 20140W-‐04. Amersfoort.
16
Appendix: Screenshots Go Pro
ID: 1
17 ID: 3
18 ID: 5
19 ID: 7
20 ID: 9
21 ID: 11
22 ID: 13
23 ID: 15
24 ID: 17
25 ID: 19
26 ID: 21
27 ID: 23
28 ID: 25
29 ID: 27
30 ID: 29
31 ID: 31
32 ID: 33
33 ID: 35
34 ID: 37
35 ID: 39
36 ID: 41
37 ID: 43
38 ID: 45
39 ID: 47
40 ID: 49
41 ID: 51
42 ID: 53
43 ID: 55
44 ID: 57
