Quantifying the distribution of submerged macrophytes in large shallow lakes
using side-scanning sonar
(NIOO-KNAW, 2017)
Name: Julia Eshuis Date: 30-06-2020
Supervised by: Elmar Becker, Harm van der Geest and Bart Schaub Bachelor thesis for Future Planet Studies, major Earth Sciences University of Amsterdam
2
Abstract
Submerged macrophytes are important for shallow lakes, as they provide food and habitat for aquatic animals. Furthermore, they play a vital role in water transparency and nutrient dynamics. Submerged macrophytes often grow in patches, which causes heterogeneity. They sometimes limit fishing and recreational activities in shallow lakes. Therefore, a clear overview of the distribution and spatial heterogeneity of macrophyte patches is required. The aims of this research are to determine the distribution of submerged macrophytes in shallow lakes, the causation of the distribution, and its effect on aquatic ecosystems. To define the distribution, narrow beam side-scanning sonar is used, that enables a single research vessel to monitor large swathes of water covered areas. The correlation between maximum vegetation densities and the distance from the shore of Lake Markermeer was measured. To determine whether macrophytes are clustered or randomly dispersed, spatial autocorrelation was calculated. With the sonar side-scanning data, higher densities are found closer to the shore, whereas smaller densities are found farther away from the shore. The distribution of the submerged macrophytes turned out to be clustered. This distribution and growth of submerged macrophytes depend on several environmental factors, such as the water depth, light intensity, and sediment type. Fish are expected to be present in low vegetation
densities, for manoeuvrability improvement, and macroinvertebrates in high vegetation densities for reduction in predation success by their predators. Spatial heterogeneity can lead to increased species diversity, as it stabilizes prey-predator interactions, by serving as a refuge for prey species. Side-scanning sonar contributes to assess the water quality of shallow lakes, supports policies in removing macrophytes, and helps to come to an agreement in the conservation of the ecosystem and maintaining recreational activities and fishing in shallow lakes.
Keywords: Macrophytes, narrow beam side-scanning sonar, spatial heterogeneity, vegetation density, shallow lakes
3
Introduction
Submerged macrophytes are important for aquatic ecosystems, as they can enhance biodiversity, improve water transparency and influence nutrient dynamics (Søndergaard et al., 2010; Van den Berg and Postema, 2001). Macrophytes are an indicator of the ecological quality of lakes, as the growth is determined by nutrient availability (Søndergaard et al., 2010). The greatest species richness will occur at intermediate nutrient levels (Bornette and Puijalon, 2011). Submerged macrophytes serve as food for aquatic animals, such as macroinvertebrates and fish, and are also consumed by aquatic birds (Van den Berg and Postema, 2001). Furthermore, they provide habitat for several aquatic animals and can serve as a refuge against predators (Parson and Matthews, 1995; Van den Berg and Postema, 2001). According to Chick and Mclvor (1997), several fish species are more abundant in vegetated than unvegetated areas. This is because, vegetated habitats have enough food resources and a reduced predation risk. Submerged macrophytes are important for the transparency of the water as they stabilise sediments, and prevent resuspension, by dampening wave action (Barko and James, 1998). Additionally, they can influence the nutrient dynamics in the water, as they take up nutrients, such as phosphorus and nitrogen, and affect oxygen conditions (Barko and James, 1998; James et al., 2004; Tarkowska-Kukuryk and Kornijów, 2008; Vonk et al., 2019).
The exact depth to where macrophytes can grow in shallow lakes depends on turbidity of the water, water temperatures and light intensity at the bottom, and therefore differs per lake and also in time (Pip, 1989; Verhofstad et al., 2017). Furthermore, high phytoplankton concentrations in the water column can reduce light availability, which is unfavourable for macrophyte growth. Also, nutrient availability influences the growth of submerged macrophytes. Too high concentrations of nutrients in the water can be disadvantage, as for example, high phosphorous concentrations can lead to domination of phytoplankton (Bornette and Puijalon, 2011). With high sediment nutrient
concentrations, however, rooted macrophytes can grow faster and taller than with less nutrient rich sediments (Bornette and Puijalon, 2011; Verhofstad et al., 2017). Additionally, stream velocities can have high influences on submerged macrophytes. The consequences depend on the hydrodynamic forces and the capacity of macrophytes to resist breakage and uprooting (Bornette and Puijalon, 2011; Schutten et al., 2005). Due to seasonal changes in submerged macrophyte cover, stream velocities can also fluctuate temporarily (Cotton, et al., 2006). In summer, due to hot temperatures, macrophytes are more abundant and cause for a decrease in stream velocity (Pip, 1989; Riis et al., 2003).
Submerged macrophytes often have a patchy distribution, because they exhibit root-clonal
propagation and therefore, cause spatial heterogeneity in the ecosystems of shallow lakes (Gantes and Caro, 2001; Li et al., 2018; Pollux et al., 2007). According to Tarkowska-Kukuryk and Kornijów (2008), a patchy distribution of submerged macrophytes facilitates water flow near sediments, creating good oxygen conditions for benthos. High densities of macrophyte patches can reduce visibility of macroinvertebrates and small fish, which reduces the capture success of their predators. However, a high density of macrophyte patches can decrease the manoeuvrability of large prey fish, which makes them more vulnerable for predators (Diehl and Kornijów, 1998; Weaver et al., 1997). A patchy distribution creates a variety of microhabitats, which are used by several organisms, such as macroinvertebrates and fish (Ferreiro et al., 2011; Pelicice et al., 2008; Weaver et al., 1997). More complex habitats may provide a greater number of individuals, especially organisms with a small body size (Ferreiro et al., 2011). However, the patchy distribution can cause habitat fragmentation and limit the dispersion capacity of several species, such as macroinvertebrates and fish, due to higher risks of predation in unvegetated areas (Diehl and Kornijów, 1998; Weaver et al., 1997).
4
Besides the ecological importance of submerged macrophytes, they can sometimes limit commercial fishing and recreational activities, such as swimming and boating, in shallow lakes, and are removed where they are deemed a nuisance (Vonk et al., 2019). To better manage submerged macrophytes in shallow lakes, a clear understanding of their spatial heterogeneity is required. Traditionally, the rake method or aerial photographs are used for quantifying macrophytes and calculating patchiness and plant density. However, the rake method can only give estimates and not exact numbers, moreover, it is labour-intensive and time-consuming (Stocks et al., 2019; Yin and Kreiling, 2011). Aerial
photographs can only distinguish submerged macrophytes that reach surface, while many macrophytes, that do not reach the surface, remain undiscovered (Marshall and Lee, 1994). An alternative method to determine the distribution of submerged macrophytes is with the use of side-scanning sonar (SSS), which is used for this research. SSS enables a single research vessel to monitor large swathes of water covered areas, using acoustic signals. SSS is less labour-intensive and time-consuming than the rake method, and is able to distinguish macrophytes that do not reach the surface (Blondel, 2010). This proposed research can help to get a clear overview of the macrophyte distribution, which will contribute to assess the water quality of shallow lakes, so appropriate policies can be made, such as the removal of submerged macrophyte. This way, an agreement can be established between parties that partly want to remove macrophytes, such as water recreants and fishermen, and parties that want to conserve the ecosystems of shallow lakes, such as aquatic ecologists, water authorities and conservationists.
The aims of this research are to determine the distribution of submerged macrophytes in shallow lakes, the causation of the distribution, and its effect on aquatic ecosystems. This leads to the following research question: ‘How can spatial heterogeneity of submerged macrophytespatchesin shallow lakes be determined?’. To answer this research question, the following three sub questions are considered:
1)‘How are submerged macrophytes distributed and to what extent can this be determined by side-scanning sonar?’. To answer this question, SSS-images were retrieved, and the spatial distribution and differences in vegetation densities of macrophytes were measured statistically using ArcGIS Pro and Matlab. Also, seasonal fluctuations of macrophyte growth were analysed.
2) ‘What causes the distribution of the submerged macrophytes?’. To answer this question, literature research was done and compared with the side-scanning sonar data of this research. 3) ‘What is the effect of the distribution of submerged macrophytes on the ecosystem of shallow lakes?’. This question will mainly be answered by literature research. A more detailed approach of the research will be discussed in the methods.
Here, it is assumed that the growth of submerged macrophytes depends on several environmental factors, such as depth, light intensity, nutrient conditions, stream velocities and water temperatures (Pip, 1989; Verhofstad et al., 2017; Vonk et al., 2019). Furthermore, it is expected that the
distribution of macrophytes is important for the ecosystem of shallow lakes, because they provide food and habitat for aquatic animals and contribute to water transparency and nutrient
concentrations in the water (Barko and James, 1998; Van den Berg et al., 2001; Vonk et al., 2019). It is hypothesized that spatial heterogeneity of macrophytes using SSS-data, can be clearly visualised and determined statistically, that submerged macrophytes have a clustered and non-randomly distribution, and that higher densities are found closer to the lakeshore.
In this research, the distribution of submerged macrophytes is determined using SSS, and testing the distribution statistically. Furthermore, it discusses the causation of the macrophyte distribution and its effect on aquatic ecosystems. Also, the discussion will explain the contribution of this thesis for the aquatic ecological research field and water management organisations.
5
Methods
Research area
This research focusses on a large research area in Lake Markermeer, which is a shallow freshwater lake located in the central Netherlands. The lake has a size of about 680 km2 and an average depth of 3.6 m (Vijverberg et al., 2011). From April till November in 2019, a monthly transect was sailed, of about 13 km in length, starting at the shore of North-Holland (Fig. 1). Due to activities blocking the sailing route, the transects from July till November where sailed south of the transects of previous months. This research will mainly focus on the transect that was sailed in July, as it is expected to see most macrophytes around this month due to high temperatures (Cotton, et al., 2006; Pip, 1989). Lake Markermeer is inhabited with abundant macrophytes, such as perfoliate and sago pondweed (Potamogeton pectinatus and Potamogeton perfoliatus) and garland (Chara aspera, Chara contraria and Chara globularis), which are all native species (Van den Berg et al., 2001; Vonk et al., 2009). The sailed transects are dominated by perfoliate pondweed. In Lake Markermeer macrophytes grow in depths around 0.4 – 3.0 m. The lake is popular for boating, water sports and fishing (Vonk et al., 2009).
Research Area with Transects in Lake Markermeer
Figure 1. Map of the research area of Lake Markermeer, showing transect locations of the research vessel Dreissena.
Literature research
To be able to answer the research question and sub questions, literature review was done on the niche of macrophytes and the effect of macrophyte patches on ecosystems of shallow lakes. Secondary data was collected, which consist of scientific articles and reports retrieved from Google Scholar. Literature research was done on factors which may affect macrophyte growth, such as water depth, nutrient availability, light intensity, stream velocities and types of substrate. To determine the ecological effects of macrophyte patches, research was done on the effect of
heterogeneity and different vegetation densities on macroinvertebrates and fish as well. In addition, the interaction between aquatic birds and macrophytes was considered. Data of other research that
6
is done on heterogeneity of submerged macrophytes, was collected and compared with the results of this research.
Side-Scanning Sonar
The SSS-data was collected with the help of the research vessel of the Institute for Biodiversity and Ecosystem Dynamics (IBED), called the R.V. Dreissena. This vessel is used for fieldwork programmes on Lake Markermeer (IBED, n.d.). The Dreissena sailed several transects from April till November, from the shore to the centre of the lake. The research vessel contains side-scanning sonar, that has transducers one on each side that can transmit frequency signals between 50 kHz and 1.2 MHz in fan-shaped pulses (Humminbird, 2019). The pulses that return first from the lake bottom
correspond to objects that are closest to the research vessel, this way differences in depth, bottom hardness and vegetation densities are seen, and an image is retrieved (Blondel, 2010). The sonar that is used for this research is a Humminbird Side Imaging sonar that uses a relatively narrow beam that can send pulses with a range up to 75 m from side to side, when using high resolutions
(Humminbird, 2019). An imaging frequency of 1275 kHz was used for this research to retrieve side-scanning sonar images.
Statistical approach
In order to better understand how the macrophytes are distributed throughout Lake Markermeer, the data were analysed and tested statistically. The SSS-images were retrieved by the sonar software Humminbird Autochart. Relative vegetation densities along the transects of the months from April till November were calculated with a built-in algorithm of the Autochart software, which means that exact calculations remain unknown. Maps with a data interpolation limit of 250 m in Autochart were made of the vegetation densities, and the seasonal fluctuation was analysed by looking at changes in macrophyte cover. Also, a map of the water depth of the transect of July was made and analysed.
The statistical approach focusses on the transect of July. The map showing relative vegetation densities of the transect of July was transferred into ArcGIS Pro, which is a geographical information system, that allows to map and edit spatial data and can perform statistical tests. The relative densities were transferred into 211-point features, and the distances between the points and the lakeshore were measured. Hundred and ten data points were randomly selected and plotted against the distance from the lakeshore using the software Matlab (which is commonly used for statistical and mathematical approaches). After this, the maximum densities per 250 m were selected and the Spearman rank correlation coefficient (with values between -1 and 1) between the maximum vegetation densities and the distance from the lakeshore was measured. A confidence interval of 95% was used, with a null hypothesis of no correlation and an alternative hypothesis indicating correlation. A high correlation means that a strong relationship between maximum densities and distance from the shore is found. A negative linear relationship was assumed between the maximum densities per 250 m and the distance from the lakeshore.
To determine whether submerged macrophytes are clustered, randomly or uniformly distributed, spatial autocorrelation of the point features was measured. This was measured with the Spatial Autocorrelation tool in ArcGIS, that calculates the Global Moran’s Index (I), a z-score and a p-value. Moran’s I can take values from -1 to 1. A value between 0 and 1, means that the distribution is considered clustered, meaning that a patchy pattern is observed, which can be caused by, for instance, environmental factors. When I is 0 the data are randomly distributed, indicating that there is no predictable pattern to explain the distribution. A value of I between -1 and 0 indicates for uniformly dispersed data, which means that the data are evenly spaced throughout the research area. Moran’s I is calculated with the following equation:
7 Eq. 1
Where n is the number of points or areas, Xi and Xj represent the value of the data for point or area of i and j, and Wij represents the geographical relationship between all pairs of point or areas of i and
j (Burt et al., 2009). After calculating the Moran’s I from the observed values, the Expected Index value was calculated. This was compared with the observed index values, and the z-score and p-value were computed, which indicates whether the difference is significant or not (ESRI, n.d.). For this research, the null hypothesis states that the macrophytes are randomly distributed throughout Lake Markermeer. As it is expected that the macrophytes are clustered, a one-sided test was used. This leads to the alternative hypothesis, that states that the macrophytes are not randomly
distributed, but clustered. The value of the z-score indicates which confidence interval is used. To prevent potential errors when calculating the spatial autocorrelation in ArcGIS, the
conceptualization of spatial relationships was set to inverse distance squared, which increases the influence between neighbouring points and decreases the influence between points that are situated far away from each other (Table 3 in appendix 5; ESRI, n.d.). This setting is favoured for the macrophyte data, because the large distances between the points can make the calculations less reliable. Additionally, the row standardization method was used, which can be implied when the distribution of the data is potentially biased due to the sampling design (Table 3 in appendix 5; ESRI, n.d.). This is applicable for this research, because the research area only shows a horizontal line, while often spatial autocorrelation is applied for more spatially distributed data (Burt et al., 2009).
The results of the statistical approach indicate how submerged macrophytes are heterogeneously distributed throughout Lake Markermeer. This helps to give answer to how they can influence ecosystems of shallow lakes, and to estimate the distribution of macrophytes in shallow lakes, other than Lake Markermeer.
Results
To determine the distribution of submerged macrophytes, monthly transects were sailed. Using SSS, graphical representations of submerged vegetation were retrieved (Fig. 2; Fig. 1 in appendix 1), that give a clear overview of the presence of vegetation in the transect of July. Also, maps with
vegetation densities of the transects from April till November were made, showing seasonal changes in macrophyte cover (Fig. 3; Fig. 1 in appendix 2). From the transect of July, a map of the water depth was made as well (Fig. 5). The vegetation densities from the transect of July were transferred into 211-point features (Fig. 4), and were used to determine the distribution of submerged
macrophytes statistically, with testing the correlation between maximum vegetation densities and the distance from the lakeshore (Fig. 6) and measuring spatial autocorrelation (Fig. 1 in appendix 5).
Imaging of submerged vegetation
Sonar images of the transect of July, at different distances from the lakeshore (Fig. 2), clearly show the presence of submerged macrophytes and individual stems are discernible. Figure 2a shows the first recognisable large patch, and figure 2d shows the last macrophytes observed in the transect. The patchiness of the macrophytes can already be distinguished, the patches closer to the lakeshore (Fig. 2a and 2b) are larger than farther away from the shore (Fig. 2c and 2d). Furthermore, the water depth of the transect is clearly visible. When sailing from the shore to the centre of the lake, it is seen that the water depth increases.
8 a)
b)
c)
d)
Figure 2. Sonar images of the transect in July, showing differences in vegetation cover at different distances from the lakeshore of Lake Markermeer (a = 200 m, b = 750 m, c = 2000 m and d = 4250 m). The water depth is indicated by the pink lines.
Vegetation densities
The transect of the different months, clearly show seasonal fluctuation of macrophyte cover (Fig. 3). Relative vegetation densities of 2 or higher are recognisable and are seen at distances smaller than 3000 m from the lakeshore. Highest densities are found in summer (Fig. 1 in appendix 2). In April, the vegetation is only recognisable near the shore (Fig. 3a). In November (Fig. 3d) almost no vegetation densities are recognisable anymore and the distribution looks less clustered, when comparing it with the transect of July (Fig. 3b).
9 a)
b)
c)
d)
Figure 3. Seasonal changes of relative vegetation densities (green) of the transects in Lake Markermeer till about 3680 m starting from the lakeshore (a = April, b = July, c = September, d = November).
In July, the highest densities are mostly found closer to the lakeshore, while farther away, only low densities are distinguishable (Fig. 4). Closer to the shore, macrophytes look more clustered than at larger distances from the shore. However, very close to the shore, till about the first 125 m, only relatively low densities are found.
V
egetation Densities in Transect of July in Lake Markermeer
10 Water depth
Figure 2 and 5 show that the water depth increases, when sailing from the shore to the centre of the lake. When looking at figure 2, the water depth increases from about 2 m (Fig. 2a) till a water depth of more than 3 m (Fig. 2d). Figure 5 shows that the waterdepth to where macrophytes are still present is between 1 m and 3 m (Fig. 3b and 5). The highest densities are found in water depths approximately between 1.75 m and 2.25 m.
Figure 5. Waterdepth of the transect of July in Lake Markermeer till about 3680 m from the lakeshore. Statisical results
To measure the distribution of submerged macrophytes and their densities statistically, the correlation coefficient and spatial autocorrelation were calculated. From all the points features, hundred and ten points were selected and plotted against the distance from the lakeshore (Fig. 6; Table 1 in appendix 3). The highest densities are observed closer to the shore (Fig. 6a). The
maximum vegetation densities that were found per 250 m, show a negative linear relationship with the formula: y = -0.0054x + 17.72 (Fig. 6b). For the maximum vegetation densities, a correlation coefficient of -0.87 with a p-value of 2.05e-04 was found, meaning that the null hypothesis is rejected (Table 1 in appendix 5; Appendix 4). Indicating that the highest macrophyte densities are found close to the shore.
a) b)
Figure 6. Scatter plots of the relationship between (maximum) vegetation densities and the distance from the lakeshore.
The outcome of the spatial autocorrelation gave a Moran’s Index of 0.45, a z-value of 9.71, and a p-value of 0.00, indicating that macrophytes have a clustered distribution (Table 2 in appendix 5). Due to the high z-score, a confidence interval of 99% is used, meaning that the null hypothesis is
rejected, and that it is less than 1 percent likely that the pattern is caused by random chance (Fig. 1 in appendix 5).
11
The results clearly show the distribution of the submerged macrophytes in Lake Markermeer, both by showing the presence and vegetation densities, and by the outcome of statistical tests. Higher densities are often found closer to shore, and macrophytes have a clustered distribution. Knowledge about the distribution of macrophytes and the water depth is required to be able to manage
macrophytes, and it helps to estimate the presence of macrophytes outside of the transect in Lake Markermeer or other shallow lakes.
Discussion
Distribution of macrophytes determined by SSS
The results show that narrow beam SSS can give a clear overview of the distribution of submerged macrophytes in shallow lakes, as it shows where macrophytes are present and where higher densities can be found. Figure 2 shows that macrophytes are often observed in patches, with larger patches closer to the lakeshore. Figure 3 indicates that macrophyte densities are highest in July, and lowest in April and November. In July, clustering of macrophytes is clearly seen, while in November and April no clear pattern is recognisable. On the sonar images of figure 2, macrophytes are still recognisable till 4250 m from the lakeshore, while on the figure 3b they are recognisable till about 3000 m. This is due to the lower vegetation density limit of 2, that is visible on the maps (Fig. 3). The calculation of the correlation coefficient shows a high negative correlation between the maximum vegetation densities and the distance from the shore, indicating that it is more likely to find high densities closer to the lakeshore (Fig. 6). However, the results also show that around the first 125 m from the shore, only low densities are present (Fig. 3 and 6), which might be caused by low water depths of about 1 m (Fig. 5). The spatial autocorrelation calculation indicates that macrophytes have a clustered distribution (Fig. 1 in appendix 5). The clustering means that the same densities are likely found with neighbouring points, indicating spatial heterogeneity (Burt et al., 2009). This can for instance be seen in figure 4, as farther away from the shore, only small densities are recognisable. Closer to the shore both high and low densities are observed (Fig. 6a). The results are in line with previous research that found spatial heterogeneity of submerged macrophytes, due to a patchy pattern and higher vegetation densities closer to the lakeshore (Ferreiro et al., 2011; Gantes and Caro, 2001; Schiemer and Prosser, 1976).
SSS is a suitable method to define the distribution of submerged macrophytes. For instance, it is more efficient than aerial photographs and the rake method. Aerial photographs only distinguish submerged macrophytes that reach the water surface (Marshall and Lee, 1994). This contrasts with SSS, that does recognise macrophytes that do not reach the surface (Fig. 2), however, aerial
photographs can cover a larger area of lakes, while especially for narrow beam SSS this would take more time. The rake method is more time-consuming, labour-intensive and less precise than SSS (Stocks et al., 2019; Yin and Kreiling, 2011). Furthermore, side-scanning sonars have also been used in other research to successful illustrate the spatial distribution of submerged macrophytes (Kruss et al., 2006; Papakonstantinou et al., 2019).
Environmental factors affecting macrophyte growth
The observed distribution of submerged macrophytes in this study is caused by several
environmental factors, such as depth, light intensity, nutrient availability and the type of substrate (Verhofstad et al., 2017; Vonk et al., 2019). When comparing figure 3b with figure 5, it is seen that macrophytes in July are observed in depths between 1 m and 3 m, and the highest densities are seen in depths between 1.75 m and 2.25 m. However, while looking at figure 2d, macrophytes are still visible at depths of more than 3 m. This difference is also caused by the lower vegetation density, which is a relative density of 2. Macrophyte patches that display lower densities might not be visible (Fig. 3b). According to Vonk et al. (2019), submerged macrophytes in Lake Markermeer grow in depths of 0.4 – 3.0 m. This partly corresponds to the results of this research. Nonetheless, it has to
12
be considered that the favourable depth for macrophytes to grow, depends on light intensity and differs per macrophyte species (Middelboe and Markager, 1997; Verhofstad et al., 2017). Light availability can be reduced by high phytoplankton growth in the water column, which is related to high phosphorous concentrations. Indicating that too high nutrient levels can be disadvantageous for macrophyte growth. The greatest species richness will occur at intermediate nutrient levels in the water column (Bornette and Puijalon, 2011). Since approximately 1980, phytoplankton biomass has decreased drastically, due to reduction of phosphorous concentrations in Lake Markermeer.
Consequently, macrophyte cover and vegetation densities increased, because of high light intensities (Noordhuis, 2010). Most macrophyte species favour fine mineral substrate to anchor themselves and grow faster and taller with higher sediment nutrient concentrations (Bornette and Puijalon, 2011; Verhofstad et al., 2017). Frequently, such substrates are highly cohesive, which favours shallow-rooted species, and prevent uprooting during high hydrodynamic forces (Bornette and Puijalon, 2011; Schutten et al., 2005). Generally, fine particle type of sediments (muds and clays) have a higher nutrient availability than coarse sediments, and therefore macrophytes often grow on these type of substrates (Jupp and Spence, 1977). This also applies for the macrophytes in Lake
Markermeer, as figure 7 shows that on the locations of the transects of April till June (Fig. 1 green line), only clay is present, and on the transects of July till November, both clay and loam are observed (Fig. 1 blue line). The patchy distribution pattern of the macrophytes is caused by root-clonal propagation, and the large distances between the patches are likely caused by spreading of seeds by aquatic birds (Gantes and Caro, 2001; Li et al., 2018; Pollux et al., 2007; Vonk et al., 2019).
Figure 7. Different soil types in Lake Markermeer (blue = clay, green = loam, yellow = loamy sand, pink = sand) (van Ledden et al., 2006).
Biotic and abiotic interactions shaped by macrophyte distribution
Defining the distribution of submerged macrophytes can be important as they influence aquatic ecosystems in multiple ways. According to previous research, the patchy macrophyte distribution is important for aquatic animals, such as macroinvertebrates, fish and aquatic birds, as they provide for habitat and food (Parson and Matthews, 1995; Van den Berg and Postema, 2001). Therefore, the distribution of macrophytes also defines the distribution of aquatic animals. Macroinvertebrates, epiphytes and bacteria use macrophytes as a substrate, they can utilize a large portion of the water column (Carpenter and Lodge, 1986; Parsons and Matthews, 1995). According to Tarkowska-Kukuryk and Kornijów (2008), a patchy distribution of submerged macrophytes facilitates water flow near sediments, creating good oxygen conditions for benthos. Several prey fish species are more abundant in vegetated than unvegetated areas. This is because vegetated habitats have enough food resources and a reduced predation risk (Chick and McIvor, 1997). High densities of macrophyte patches can reduce visibility of macroinvertebrates and small prey fish, which reduces the capture
13
success of their predators. However, a high density of macrophyte patches can decrease the manoeuvrability of large prey fish, which makes them more vulnerable for predators (Chick and McIvor, 1997; Diehl and Kornijów, 1998; Weaver et al., 1997). Therefore, it is expected to observe large prey fish in less dense macrophyte patches and macroinvertebrates and small prey fish in denser patches (Fig. 5; Pelicice et al., 2008). Predatory fish are likely to wait near the periphery of vegetated areas and attack when prey becomes visible (Heck and Orth, 1980). A patchy distribution creates a variety of microhabitats, which are inhabited by several aquatic animals, such as
macroinvertebrates and fish (Ferreiro et al., 2011; Pelicice et al., 2008; Weaver et al., 1997). A higher species diversity can be caused by heterogeneity through the provisioning of niches and diverse environmental resources (Cornacchia et al., 2018; Ferreiro et al., 2011). Spatial heterogeneity can lead to increased species diversity within a community, because it stabilizes prey-predator interactions, by serving as a refuge for prey species (Gilinksy, 1984). However, the patchy
distribution can cause habitat fragmentation and limit the dispersion capacity of several species, due to higher risks of predation in unvegetated areas (Diehl and Kornijów, 1998; Weaver et al., 1997).
Besides the influence on aquatic animals, macrophyte patches can affect abiotic factors as well. Submerged macrophyte patches are important for water transparency, as they stabilise sediments, and prevent resuspension, by dampening wave action (James et al., 2004; Penning et al., 2009). Furthermore, dense patches of submerged macrophytes can alter flow regimes, which reduces stream velocity and leads to high water levels (Bal et al., 2001; Champion and Tanner, 2000; Clarke, 2002; Penning et al., 2009). Seasonal fluctuations of macrophyte cover have to be considered, as in summer higher percentages of macrophyte cover are found than in winter (Fig. 3; Cotton et al., 2006; Riis et al., 2003). Macrophytes can prevent erosion of lake sediments, especially high
vegetation densities (Bouma et al., 2009). Additionally, they can influence nutrient dynamics in the water, as they take up nutrients, such as phosphorus and nitrogen, and regulate oxygen fluxes (Barko and James, 1998; Collier et al., 1999). What is more, 1 to 10% of the photosynthetically-fixed carbon of actively growing macrophytes is released in the water as dissolved organic compounds (Carpenter and Lodge, 1986). These compounds contribute to the metabolism of bacteria and epiphytes. Generally, macrophyte patches can serve as nutrient sinks for particulate matter and as sources for organic carbon (Carpenter and Lodge, 1986).
Experimental considerations
Altogether, the results show that SSS gives a clear overview of the distribution of submerged macrophytes in Lake Markermeer. Furthermore, SSS can help to estimate the distribution of
submerged macrophytes in shallow lakes, other than Lake Markermeer, and can help governmental organisations, that take care of water management, remove macrophytes in proper areas. This way, recreational activities and fishing can be proceeded and damage of the ecosystem in lakes is
minimized. When understanding the distribution of macrophytes, the distributions of aquatic animals can be predicted as well. Therefore, SSS is used by fishermen to predict the presence of fish and also to detect certain vegetation densities (Pelicice et al., 2008). Additionally, it can help aquatic ecologists to do research on aquatic ecosystems, and conserving and managing of shallow lakes can be improved, by finding the right balance between recreational activities and maintaining the ecosystem in sustainable ways.
Some uncertainties about this research have to be considered. Firstly, the distribution of the
macrophytes is only known from a straight horizontal line of the transect. This might have influenced the outcome of the spatial autocorrelation test, as it not known how the macrophyte distribution looks like outside of the transect. The transect shows a more one-dimensional distribution, while often spatial autocorrelation is calculated for more spatial areas (Burt et al., 2009). However, these uncertainties have been reduced as much as possible, due to the settings of the Spatial
14
have sailed zigzagging through a large area in the lake, instead of one straight line, however, this would be more time-consuming. Secondly, it has to be considered that the transects from April till June were sailed north of the transects from July till November, meaning that the transects might have looked different at the location more south. Furthermore, an important shortcoming is that the relative densities were calculated by an algorithm created by the Humminbird Autochart software. The calculations of this algorithm are unknown, and it might have caused potential errors. Lastly, the maps of figures 3 and 4 were interpolated with 250 m, meaning the maps are mainly estimates of how the transects would look like. Another possibility is to use sonars that scan wider ranges.
Further research should focus on determining the correlation coefficient and spatial autocorrelation of other months, to get to know more about the exact seasonal changes in distribution of
submerged macrophytes. Additionally, research can be done about the relationship between different vegetation densities and number of fish and macroinvertebrates. It will be interesting to see if the assumption will be met, that large prey fish mostly occur in low vegetation densities and macroinvertebrates and small prey fish in high densities.
Conclusion
In answering the research question, which states how spatial heterogeneity of submerged
macrophyte patches in shallow lakes can be determined, it can be concluded that narrow beam SSS is a suitable method for quantifying the distribution of submerged macrophytes. It is more efficient than the rake method and aerial photographs, and it can clearly show the presence of macrophytes and the vegetation densities. Higher densities are found closer to the lakeshore, whereas smaller densities are found farther away from the shore in Lake Markermeer. The distribution of the submerged macrophytes displayed a clustered spatial distribution. The patchy distribution of submerged macrophytes depend on several environmental factors, such as the water depth, light intensity, and the type of the substrate. Also, root-clonal propagation and the further spreading of seeds by aquatic birds may play a role. Submerged macrophytes are important for aquatic animals, as they provide food and habitat. Large prey fish are expected to be present in low vegetation densities, where they have increased manoeuvrability, and macroinvertebrates and small prey fish in high vegetation densities, to reduce predation success of their predators. Spatial heterogeneity can lead to increased species diversity, as it stabilizes prey-predator interactions, by serving as a refuge for prey species. However, the patchy distribution can cause habitat fragmentation and limit the dispersion capacity of several species. Submerged macrophyte patches can affect abiotic factors as well. For instance, they cause water transparency, and higher vegetation densities prevent erosion and cause lower stream velocities. Additionally, they influence the nutrient dynamics in the water. SSS can help to estimate the distribution of submerged macrophytes in shallow lakes, and
contributes to inform policy decisions about water management, such as the removal of
macrophytes. Also, SSS is used to predict presence of fish by fishermen, and to find the right balance between maintaining recreational activities and fishing in shallow lakes and managing the ecosystem in sustainable ways.
Acknowledgements
I want to thank the University of Amsterdam and the Institution of Biodiversity and Ecosystem Dynamics for funding this research and providing data. Particularly, appreciation is given to Elmar Becker, Harm van der Geest, Bart Schaub and Brous Ernst (Water Authority Rijnland) for advising me in making decisions for this project and helping to set up this research.
15
References
Bal, K., Struyf, E., Vereecken, H., Viaene, P., De Doncker, L., de Deckere, E., ... & Meire, P. (2011). How do macrophyte distribution patterns affect hydraulic resistances?. Ecological
engineering, 37(3), 529-533.
Barko, J. W., & James, W. F. (1998). Effects of submerged aquatic macrophytes on nutrient dynamics, sedimentation, and resuspension. In The structuring role of submerged macrophytes in lakes (pp. 197-214). Springer, New York, NY.
Blondel, P. (2010). The handbook of sidescan sonar. Chichester, UK: Springer Science & Business Media.
Bornette, G., & Puijalon, S. (2011). Response of aquatic plants to abiotic factors: a review. Aquatic Sciences, 73(1), 1-14.
Bouma, T. J., Friedrichs, M., Van Wesenbeeck, B. K., Temmerman, S., Graf, G., & Herman, P. M. J. (2009). Density‐dependent linkage of scale‐dependent feedbacks: A flume study on the intertidal macrophyte Spartina anglica. Oikos, 118(2), 260-268.
Burt, J. E., Barber, G. M., & Rigby, D. L. (2009). Elementary Statisitics for Geopgraphers (3rd edition). New York, NY: The Guilford Press.
Carpenter, S. R., & Lodge, D. M. (1986). Effects of submersed macrophytes on ecosystem processes. Aquatic botany, 26, 341-370.
Champion, P. D., & Tanner, C. C. (2000). Seasonality of macrophytes and interaction with flow in a New Zealand lowland stream. Hydrobiologia, 441(1), 1-12.
Chick, J. H., & Mclvor, C. C. (1997). Habitat selection by three littoral zone fishes: effects of predation pressure, plant density and macrophyte type. Ecology of Freshwater Fish, 6(1), 27-35.
Clarke, S. J. (2002). Vegetation growth in rivers: influences upon sediment and nutrient dynamics. Progress in Physical Geography, 26(2), 159-172.
Collier, K. J., Champion, P. D., & Croker, G. F. (1999). Patch-and reach-scale dynamics of a
macrophyte-invertebrate system in a New Zealand lowland stream. Hydrobiologia, 392(2), 89-97.
Cotton, J. A., Wharton, G., Bass, J. A. B., Heppell, C. M., & Wotton, R. S. (2006). The effects of seasonal changes to in-stream vegetation cover on patterns of flow and accumulation of sediment. Geomorphology, 77(3-4), 320-334.
Cornacchia, L., Van De Koppel, J., van der Wal, D., Wharton, G., Puijalon, S., & Bouma, T. J. (2018). Landscapes of facilitation: how self‐organized patchiness of aquatic macrophytes promotes diversity in streams. Ecology, 99(4), 832-847.
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.
16
ESRI. (n.d.). Spatial Autocorrelation (Global Moran’s I) (Spatial Statistics). Retrieved on 7 June 2020, from:
https://pro.arcgis.com/en/pro-app/tool-reference/spatial-statistics/spatial-autocorrelation.htm
Ferreiro, N., Feijoó, C., Giorgi, A., & Leggieri, L. (2011). Effects of macrophyte heterogeneity and food availability on structural parameters of the macroinvertebrate community in a Pampean stream. Hydrobiologia, 664(1), 199-211.
Gantes, H. P., & Caro, A. S. (2001). Environmental heterogeneity and spatial distribution of macrophytes in plain streams. Aquatic Botany, 70(3), 225-236.
Gilinsky, E. (1984). The role of fish predation and spatial heterogeneity in determining benthic community structure. Ecology, 65(2), 455-468.
Heck, Jr. K. L., & Orth, R. J. (1980). Seagrass habitats: the roles of habitat complexity, competition and predation in structuring associated fish and motile macroinvertebrate assemblages. In Estuarine perspectives (pp. 449-464). Academic Press.
Humminbird. (2019). Solix: Operations Manual. Retrieved on 7 June 2020, from:
https://humminbird.johnsonoutdoors.com/sites/johnsonoutdoors-store/files/assets/misc/HU/H/HUM_productmanual_OPS_SOLIX_SRS_532566E/HUM_productmanu al_OPS_SOLIX_SRS_532566E.pdf
IBED. (n.d.). Reasearch vessel RV Dreissena. Retrieved on 12 April 2020, from:
https://ibed.uva.nl/facilities/fieldwork-facilities/research-vessel-rv-dreissena/research-vessel-rv-dreissena.html?cb
James, W. F., Barko, J. W., & Butler, M. G. (2004). Shear stress and sediment resuspension in relation to submersed macrophyte biomass. Hydrobiologia, 515(1-3), 181-191.
Jupp, B. J., & Spence, D. H. N. (1977). Limitations of macrophytes in a eutrophic lake, Loch Leven: II. Wave action, sediments and waterfowl grazing. The Journal of Ecology, 431-446.
Kruss, A., Tęgowski, J., Wiktor, J., & Tatarek, A. (2006). Acoustic estimation of macrophytes in the Hornsund fjord (the Svalbard Archipelago). Hydroacoustics, 9, 89-96.
Li, W., Li, Y., Zhong, J., Fu, H., Tu, J., & Fan, H. (2018). Submerged macrophytes exhibit different phosphorus stoichiometric homeostasis. Frontiers in plant science, 9, 1207.
Marshall, T. R., & Lee, P. F. (1994). Mapping aquatic macrophytes through digital image analysis of aerial photographs: an assessment. Journal of Aquatic Plant Management, 32, 61-66.
Middelboe, A. L., & Markager, S. (1997). Depth limits and minimum light requirements of freshwater macrophytes. Freshwater Biology, 37(3), 553-568.
NIOO-KNAW. (2017, 21 December). Woekerende waterplanten: maaien of niet? Retrieved on 14 June, 2020 from: https://www.naturetoday.com/intl/nl/nature-reports/message/?msg=24019
Noordhuis, R. (2010). Ecosysteem IJsselmeergebied: nog altijd in ontwikkeling. Retrieved on 7 June 2020, from: http://publicaties.minienm.nl/documenten/ecosysteem-ijsselmeergebied-nog-altijd-in-ontwikkeling-trends-en
17
Papakonstantinou, M., Papatheodorou, G., Christodoulou, D., & Papastergiadou, (2019). E. Seasonal Variations of Water Quality and Aquatic Macrophytes Survey based on Side Scan Sonar, in a Shallow Coastal Lagoon (Gialova, SW Peloponnesus, Greece).
Parsons, J. K., & Matthews, R. A. (1995). Analysis of the associations between macroinvertebrates and macrophytes in a freshwater pond. Northwest Science, 69(4).
Pelicice, F. M., Thomaz, S. M., & Agostinho, A. A. (2008). Simple relationships to predict attributes of fish assemblages in patches of submerged macrophytes. Neotropical Ichthyology, 6(4), 543-550.
Penning, W. E., Raghuraj, R., & Mynett, A. E. (2009). The effects of macrophyte morphology and patch density on wave attenuation.
Pip, E. (1989). Water temperature and freshwater macrophyte distribution. Aquatic Botany, 34(4), 367-373.
Pollux, B. J. A., Jong, M. D. E., Steegh, A., Verbruggen, E., Van Groenendael, J. M., & Ouborg, N. J. (2007). Reproductive strategy, clonal structure and genetic diversity in populations of the aquatic macrophyte Sparganium emersum in river systems. Molecular Ecology, 16(2), 313-325.
Riis, T., Biggs, B. J., & Flanagan, M. (2003). Seasonal changes in macrophyte biomass in South Island lowland streams, New Zealand. New Zealand Journal of Marine and Freshwater Research, 37(2).
Schiemer, F., & Prosser, M. (1976). Distribution and biomass of submerged macrophytes in Neusiedlersee. Aquatic Botany, 2, 289-307.
Schutten, J., Dainty, J., & Davy, A. J. (2005). Root anchorage and its significance for submerged plants in shallow lakes. Journal of Ecology, 93(3), 556-571.
Søndergaard, M., Johansson, L. S., Lauridsen, T. L., JØRGENSEN, T. B., Liboriussen, L., & Jeppesen, E. (2010). Submerged macrophytes as indicators of the ecological quality of lakes. Freshwater Biology, 55(4), 893-908.
Stocks, J. R., Rodgers, M. P., Pera, J. B., & Gilligan, D. M. (2019). Monitoring aquatic plants: An evaluation of hydroacoustic, on-site digitising and airborne remote sensing techniques. Knowledge & Management of Aquatic Ecosystems, 420, 27.
Tarkowska-Kukuryk, M., & Kornijów, R. (2008). Influence of spatial distribution of submerged macrophytes on Chironomidae assemblages in shallow lakes. Polish Journal of Ecology, 56(4), 569-579.
Van den Berg, M. S., Doef, R. W., & Postema, J. (2001). Waterplanten in het IJsselmeergebied. De Levende Natuur, 102(5), 237-241.
Van Ledden, M., Gerrits, G. W. R., van Kessel, T. & Mosselman, E. (2006). Verdiepingsslag en maatregelen slibproblematiek Markermeer–Analyse kennisleemten en inventarisatie maatregelen. Royal Haskoning and WL Delft Hydraulics.
Verhofstad, M. J. J. M., Núñez, M. A., Reichman, E. P., van Donk, E., Lamers, L. P. M., & Bakker, E. S. (2017). Mass development of monospecific submerged macrophyte vegetation after the restoration
18
of shallow lakes: roles of light, sediment nutrient levels, and propagule density. Aquatic Botany, 141, 29-38.
Vijverberg, T., Winterwerp, J. C., Aarninkhof, S. G. J., & Drost, H. (2011). Fine sediment dynamics in a shallow lake and implication for design of hydraulic works. Ocean Dynamics, 61(2-3), 187-202.
Vonk, A. J., Verhofstad, M. J. J. M., & van der Geest, H. G. (2019). Ondergedoken waterplanten in het Markermeer: vragen en antwoorden. Retrieved on 12 April 2020, from:
https://www.groot-waterland.nl/wp-content/uploads/2019/12/waterplanten-markermeer-uva-rapport-2019.pdf
Weaver, M. J., Magnuson, J. J., & Clayton, M. K. (1997). Distribution of littoral fishes in structurally complex macrophytes. Canadian Journal of Fisheries and Aquatic Sciences, 54(10), 2277-2289.
Yin, Y., & Kreiling, R. M. (2011). The evaluation of a rake method to quantify submersed vegetation in the Upper Mississippi River. Hydrobiologia, 675(1), 187.
19
Appendix 1 – Side-scanning sonar images
a) b) c) d) e) f)
Figure 1. Sonar images of the transect in July, showing differences in vegetation cover at different distances from the lakeshore of Lake Markermeer (a = 200 m, b = 500 m, c = 750 m, d = 1000 m, e = 2000 m). The water depth is indicated by the pink lines (figure continues on next page).
20 g)
h)
i)
Figure 1. Sonar images of the transect in July, showing differences in vegetation cover at different distances from the lakeshore of Lake Markermeer (g = 3250 m, h = 3500 m, i = 4250 m). The water depth is indicated by the pink lines (figure continued from previous page).
21
Appendix 2 - Seasonal Fluctuation of Submerged Macrophytes
a) b) c) d) e) f) g) h)
Figure 1. Seasonal changes of relative vegetation densities (green) of the transects in Lake Markermeer till about 3680 m starting from the lakeshore (a = April, b = May, c= June, d = July, e = Augsutus, f = September, g = October, h = November).
22
Appendix 3 – Vegetation Densities and Distance from the Lakeshore
Table 1. Different relative densities against the distance from the lakeshore of Lake Markermeer.
23
Appendix 4 – Script in Matlab
% Script for Bachelor Project - Future Planet Studies % June 2020
% Name: Julia Eshuis
% loading data of density points load('DensityPoints2.mat') Density = DensityPoints2.Density; Distance = DensityPoints2.Distance; %plotting data figure plot(Distance, Density,'o') ylim([0 25]) xlabel('Distance (m)') ylabel('Density')
title('Vegetation Densities of Transect in July') % maximum densities per 250 m
%250
Dis0 = Distance <= 250 Dens250 = Density(Dis0) Max250 = max(Dens250) %500
Dis1 = Distance <= 500 & Distance > 250 Dens500 = Density(Dis1);
Max500 = max(Dens500); %750
Dis2 = Distance <= 750 & Distance > 500 Dens750 = Density(Dis2);
Max750 = max(Dens750); %1000
Dis3 = Distance <= 1000 & Distance > 750 Dens1000 = Density(Dis3);
Max1000 = max(Dens1000); %1250
Dis4 = Distance <= 1250 & Distance > 1000 Dens1250 = Density(Dis4);
Max1250 = max(Dens1250); %1500
Dis5 = Distance <= 1500 & Distance > 1250 Dens1500 = Density(Dis5);
Max1500 = max(Dens1500); %1750
Dis6 = Distance <= 1750 & Distance > 1500 Dens1750 = Density(Dis6);
Max1750 = max(Dens1750); %2000
Dis7 = Distance <= 2000 & Distance > 1750 Dens2000 = Density(Dis7)
Max2000 = max(Dens2000) %2250
Dis8 = Distance <= 2250 & Distance > 2000 Dens2250 = Density(Dis8)
Max2250 = max(Dens2250) %2500
Dis9 = Distance <= 2500 & Distance > 2250 Dens2500 = Density(Dis9);
Max2500 = max(Dens2500); %2750
Dis10 = Distance <= 2750 & Distance > 2500 Dens2750 = Density(Dis10);
Max2750 = max(Dens2750); %3000
Dis11 = Distance <= 3000 & Distance > 2750 Dens3000 = Density(Dis11);
24
Max3000 = max(Dens3000);
Max = [Max250; Max500; Max750; Max1000; Max1250; Max1500; Max1750; Max2000;Max2250; Max2500; Max2750; Max3000];
DistanceMean = [125;375;625;875;1125;1375;1625;1875;2125;2375;2625;2875]; %checking for normality of variables -> not normal, Spearman's rank correlation will be used
histogram(Max)
histogram(DistanceMean)
%checking for normality of residuals -> normal distribution of residuals md1 = fitlm(DistanceMean, Max)
histogram(md1.Residuals.Raw,[-8:4:8]) % checking for outliers -> no outliers
outlierdown = quantile(Max,0.25) - (1.5*iqr(Max)) outlierup = quantile(Max,0.75) + (1.5*iqr(Max))
outlierdown2 = quantile(DistanceMean,0.25) - (1.5*iqr(DistanceMean)) outlierup2 = quantile(DistanceMean,0.75) + (1.5*iqr(DistanceMean)) % fitting line and plotting data
p = polyfit(DistanceMean, Max,1); y = polyval(p,1:3000); figure plot(DistanceMean,Max,'x') hold on plot(y,'-r') ylim([0 25]) legend('Data','Trend Line') ylim([0 25]) xlabel('Distance (m)') ylabel('Density')
title('Maximum Vegetation Density per 250 m') text(2100,20.5,'y = -0.0054x + 17.72')
%correlation
25
Appendix 5 – Descriptive Statistics
Table 1. Summary of point data and output of Spearman’s rank test in Matlab.
Data Number Min Max Mean std Rho p-value
Vegetation densities 110 2 22 6.2182 4.7572 - - Maximum vegetation
densities per 250 m
12 2 22 9.6667 6.0202 -0.8736 2.0469e-04
Table 2. The output of the Spatial Autocorrelation Tool in ArcGIS.
Moran’s Index 0.450545
Expected Index -0.004762
Variance 0.002198
z-score 9.711257
p-value 0.000000
Table 3. Settings for the Spatial Autocorrelation Tool in ArcGIS.
Input Feature Class Vegetation Density
Input Field DENSITY
Conceptualization INVERSE_DISTANCE_SQUARED
Distance Method EUCLIDEAN
Row Standardization True
Distance Threshold 279.1248 Meters
Figure 1. Output from the Spatial Autocorrelation Tool in ArcGIS of the distribution of macrophytes within the transect of July, showing the significance level and critical value.
