Abstract
Abstract
Rouhe AC, Rueter JG. Reducing surface accumulation of Aphanizomenon flos-aquae using wetland water to increase cellular turgor pressure and interfere with buoyancy regulation. Lake Reserve Manage. 34:426–446.
The ability to regulate buoyancy (sinking and floating) using cellular gas vesicles is a unique characteristic that allows many common bloom-forming cyanobacteria to accumulate at water surfaces and dominate systems. Typical control and management strategies of cyanobacterial blooms include nutrient manipulation and phosphorus reduction, which are effective but do not reduce the advantage of buoyancy control. Since buoyancy control is based upon a mechanism driven by photosynthesis, along with environmental conditions that trigger vesicle formation and ion exchange, buoyancy regulation can be influenced by manipulating extracellular conditions. In this study, we manipulated extracellular conditions using wetland water and additions of potassium, sodium, and calcium in small-scale lab experiments containing Aphanizomenon flos-aquae from Upper Klamath Lake, Oregon. The results indicate a target mixture of 10% wetland water reduces surface accumulation, increases cellular turgor pressure (a measure of the ability of gas vesicle–forming cells to control buoyancy), and leads to fewer cyanobacterial filament rafts near the surface of the water column. By adding ions at the same concentration as the target wetland mixture, similar results were found. This research represents the basis of a possible strategy for mitigating surface blooms of buoyant cyanobacteria in lakes using wetland water and/or ion additions that could be used in tandem with nutrient manipulation and phosphorus reduction.
Harmful algal blooms of cyanobacteria (CyanoHABs) have become an increasing problem worldwide (Sukenik et al. 2015). CyanoHABs are primarily formed by species from the following genera: Dolichospermum (formerly the planktonic members of Anabaena), Aphanizomenon, Cylindrospermopsis, Lyngbya, Microcystis, Nodularia, Oscillatoria, Planktothrix, Woronichinia, and Trichodesmium (O’Neil et al. 2012, Paerl and Otten 2013). Species from these genera utilize a wide variety of present conditions to become the dominant taxa in systems (Dokulil and Teubner 2000, Paerl and Otten 2013). They predominate by using characteristics such as nitrogen fixation, buoyancy control, clustering of cells and filaments, toxin production, tolerance of low CO2, and superior growth in low light (Kalff and Knoechel 1978, Zevenboom and Mur 1980, Lindholm et al. 1989, Oliver 1994, Dokulil and Teubner 2000, O’Neil et al. 2012). While each cyanobacteria species employs its own unique combination of these characteristics, the major CyanoHAB species from the above genera share the ability to regulate buoyancy (floating and sinking) and use surface accumulation to gain an advantage over nonbuoyant phytoplankton taxa (Klemer and Konopka 1989, Walsby 1994, Paerl and Otten 2013).
Surface accumulation is a key advantage for the dominance of buoyancy regulating cyanobacteria (Reynolds et al. 1987) but CyanoHABs that accumulate at the surface of lakes can have adverse effects on lake ecology (Scheffer et al. 1997, Dokulil and Teubner 2000). Buoyant cyanobacteria accumulate at the surface of a water body by forming gas vesicles in individual cells (Oliver 1994). The main factors that trigger gas vesicle formation are low light and high phosphorus availability (Konopka et al. 1987, Walsby 1994). Once triggered, gas vesicles are often formed in excess, leading to dense accumulations of cyanobacteria cells, filaments, and colonies at the water surface that can persist for days, weeks, and months (Zohary and Roberts 1990, Walsby et al. 1991). A dense surface accumulation drastically reduces light availability in the lower water column (Scheffer et al. 1997) creating a low-light environment that favors cyanobacteria growth (Zevenboom and Mur 1980, Dokulil and Teubner 2000). Experimental tests of competition between phytoplankton taxa show that cyanobacteria outcompete diatoms and green algae in stable, low-light conditions (Zevenboom and Mur 1980, Huisman 1999, Huisman et al. 2004). Thus, shading from surface accumulation creates conditions that favor cyanobacteria propagation over other competing phytoplankton taxa, leading to an unhealthy system with low phytoplankton diversity, diminished zooplankton grazing, and poor water quality (Scheffer et al. 1997, Dokulil and Teubner 2000, Llames et al. 2009).
A methodology that facilitates reduction of surface accumulation could be a valuable management tool that would complement current CyanoHAB management strategies. Reducing surface accumulations of buoyancy regulating CyanoHABs would allow more light to penetrate the water column, diminishing the advantage of superior cyanobacteria growth in low light (Huisman 1999, Klausmeier and Litchman 2001). Preventing or reducing cyanobacteria surface accumulation would require conditions that lead to an increase in cell density and a reduction in gas vesicle number and effectiveness. This shift can be accomplished by establishing conditions that lead to an increase in cellular turgor pressure (Oliver and Walsby 1984, Thomas and Walsby 1985). Turgor pressure in cells of buoyant cyanobacteria naturally develops as cells accumulate carbohydrates and exchange ions during photosynthesis (Oliver 1994). As turgor pressure increases, gas vesicles begin to collapse under the pressure (Dinsdale and Walsby 1972, Kinsman et al. 1991). This process is considered ubiquitous among gas vesicle–forming cyanobacteria (Oliver 1994) and has been shown to play a major role in buoyancy regulation of Aphanizomenon, Dolichospermum, Gloeotrichia, and Microcystis (Oliver and Walsby 1984, Thomas and Walsby 1985, Kromkamp et al. 1986, Oliver 1994, Walsby 1994).
A crucial aspect of the management of buoyant CyanoHABs, by means of increasing cellular turgor pressure in order to collapse gas vesicles, is the ability to induce and accelerate cellular turgor pressure. Studies by Anthony Walsby throughout the 1970s and 1980s conclusively demonstrated the ability to induce and accelerate cellular turgor pressure and gas vesicle collapse in cyanobacteria. Walsby developed an instrument that applied incremental external pressure to a small chamber containing cyanobacteria and measured vesicle collapse from changes in the refractive index (Walsby 1971). By changing the conditions under which cyanobacteria were grown or preconditioned, this instrument allowed for the calculation and comparison of turgor pressure and gas vesicle changes related to preconditioning. Specifically, it was shown that increases in light intensity (Walsby 1971, Oliver and Walsby 1984) and duration of light exposure (Grant and Walsby 1977, Thomas and Walsby 1985) led to an increase in cellular turgor pressure in Microsystis and Dolichospermum. The relationships between light duration, light intensity, and turgor pressure were also shown for Aphanizomenon flos-aquae by Kromkamp et al. (1986) using a similar method. Physical and chemical changes during preconditioning also caused induction and acceleration of cellular turgor pressure and gas vesicle collapse (Walsby 1971). Addition of potassium ions led to an increase in turgor pressure and vesicle collapse (Allison and Walsby 1981), while addition of a hypertonic sucrose solution led to a decrease in turgor pressure and vesicle collapse (Walsby 1971, Dinsdale and Walsby 1972, Grant and Walsby 1977, Thomas and Walsby 1985). Thus, cellular turgor pressure in buoyant CyanoHABs can be manipulated by altering water chemistry, light exposure, and light intensity, making the disruption of buoyancy regulation and the reduction of surface accumulation a viable strategy. While induction of turgor pressure by manipulating light intensity and/or duration of light exposure is not a practical management strategy for lakes, altering water chemistry conditions through the manipulation of ions or the addition of other substances that increase cellular turgor pressure may be feasible for lake-wide management.
In this study, we investigated conditions that increase cellular turgor pressure and influence the buoyancy and surface accumulation of Aphanizomenon flos-aquae collected from Upper Klamath Lake (UKL), Oregon, using small-scale, short-term laboratory trials. Our main focus was to understand the effect of humic-rich wetland water on the buoyancy and surface accumulation of A. flos-aquae. Our preliminary studies on the impact of adding humic-rich wetland water to lake water (data not included here) showed a decrease in buoyancy control and surface accumulation when rafts of A. flos-aquae from UKL were exposed to water from an adjacent wetland. While the idea of measuring the effect of wetland water on cellular turgor pressure and surface accumulation is novel for this system, wetland water has long been suspected of adversely affecting cyanobacteria in UKL. Other studies of the area have noted a lack of A. flos-aquae dominance in and around wetlands connected to the lake (Geiger et al. 2005), and small-scale experiments have indicated a link between humic conditions and a reduction in A. flos-aquae from UKL (Haggard et al. 2013). We also investigated the effect of potassium, sodium, and calcium ion additions. These 3 cations were selected based upon their involvement in the mechanisms for cation exchange during photosynthesis (Kaplan et al. 1989) and previous studies that have linked them to the regulation of buoyancy in cyanobacteria (Allison and Walsby 1981, Reed and Walsby 1985, Wang et al. 2011).
To compare the turgor pressure between treatments, we developed a method based upon Walsby (1971), but used a commercially available pressure chamber for external pressure application and recorded the buoyancy loss of A. flos-aquae as a proxy for vesicle collapse. In this paper, we describe the adapted method and show the results of our laboratory tests that capture the same trend as Walsby’s method for manipulations of light intensity, duration of light exposure, and potassium. For the laboratory trials conducted in this study for wetland water and ions there were 2 hypotheses. First, we anticipated a direct relationship between turgor pressure and increasing treatment levels of wetland water and ions. Second, we expected surface accumulation to decrease as treatment levels of wetland water and ions increased. For each treatment type (wetland water, potassium, etc.), we expected that treatment level increase would lead to a progressive increase in turgor pressure, and a progressive decrease in surface accumulation.
Study taxa
The taxa of focus for this study was Aphanizomenon flos-aquae Ralfs ex Born. & Flah. var. flos-aquae Aph K-2 (Li et al. 2000) from UKL. A. flos-aquae is a filamentous, gas vesicle–forming, diazatrophic cyanobacterium. In UKL, A. flos-aquae filaments cluster together into large, macroscopic structures referred to as flakes, clusters, or rafts. The A. flos-aquae rafts in UKL are macroscopic, having the appearance of thin blades of grass in the lake. Rafts are commonly shorter than 0.5 cm, but large rafts can reach lengths close to 2 cm (Rouhe 2008).
Study site
UKL is part of a 2-lake system located in a volcanic rock–dominated graben on the eastern side of the Cascade Range in southern Oregon (Snyder and Morace 1997). The main water body in this system is UKL, with a surface area of 232 km2, an average depth of 2.8 m, and a maximum depth of 15.2 m. Just north of UKL is Agency Lake. Agency Lake is much smaller than UKL with a surface area of 38 km2, an average depth of 0.9 m, and a maximum depth of 3 m (Hoilman et al. 2008). Until 2007, UKL and Agency Lake were connected only by a narrow passage between the Klamath Wildlife Refuge (a large, reed-dominated wetland area) and a levee that formed the barrier of a drained wetland area. In 2007, sections of the levees around the drained wetland were removed, flooding an area of about 30 km2 (Fig. 1). Completed in 2009, this reclamation project connected UKL and Agency Lake, forming one contiguous system that consists of UKL, Agency Lake, and many littoral wetlands for a total area of 305 km2 and an average depth of 2.6 m (Eldridge et al. 2013, Wood et al. 2013). There are 2 main inflows to the system: the Wood River, which flows into the north end of Agency Lake, and the Williamson River, which flows into the north end of UKL (Fig. 1). The outflow from the system occurs through the Link River Dam located at the southern end of UKL.
Published online:
02 October 2018Figure 1. Upper Klamath Lake, Oregon, showing key lake features and study locations.
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/ulrm20/2018/ulrm20.v034.i04/10402381.2018.1466011/20190117/images/medium/ulrm_a_1466011_f0001_b.jpg"})
Figure 1. Upper Klamath Lake, Oregon, showing key lake features and study locations.
Since the formation of the lakes in the Klamath Basin over 10,000 yr ago, wetlands have been an integral part of the Upper Klamath Lake system (Dicken 1980, Snyder and Morace 1997). During the 1800s, land in the region began to be used for agriculture (Snyder and Morace 1997) and near-lake wetlands began to be used for wild hay harvest and cattle grazing (Akins 1970). By the middle of the 1970s 64% of the wetlands directly connected to UKL and Agency Lake had been drained for agriculture and livestock use (Snyder and Morace 1997). Based upon historical accounts and paleolimnological sediment cores, UKL has always been a very productive system (Cope 1883, Phinney and Peak 1961, Eilers et al. 2004). However, the system was not always dominated by A. flos-aquae. Some of the first reports of A. flos-aquae occur in the 1930s (Bonell and Mote 1942). Eilers et al. (2004) demonstrated that the system was dominated by diatoms prior to the 1800s, with increasing dominance of A. flos-aquae in the last 200 yr corresponding with increased agriculture and livestock in the region. Currently, UKL and Agency Lake are considered hypereutrophic (Kann 1998, Hoilman 2008). UKL and Agency Lake are typically dominated by surface blooms of A. flos-aquae from May to October (Kann 1998). However, the system has a strong spring diatom bloom dominated by Fragilaria (Carmichael et al. 2000) and small blooms of Dolochospermum flos-aquae, Woronichinia sp., Gloeotrichia sp., and Microcystis sp. occur with A. flos-aquae throughout the summer and fall (Eldridge et al. 2017). In recent years, toxigenic forms of Microcystis have been accumulating at the surface, most likely due to elevated levels of nitrogen from A. flos-aquae (Eldridge et al. 2017), forming scums and resulting in elevated levels of microcystin. In this system, A. flos-aquae is not toxigenic and elevated levels of microcystin have been directly linked to Microcystis (Eldridge et al. 2017).
Small-scale lab treatments
A. flos-aquae rafts were collected at the southern end of UKL, far from any influence of wetland water, on the end of Putnams Point (WGS84 42°14′20.5044″N, 121°48′27.8208″W), a rocky shore about 0.5 km north of the Link River Dam (Fig. 1). Rafts were collected in amber bottles each day of the study before sunrise to avoid the onset of photosynthesis. Lake water for each treatment was collected each morning of the study with A. flos-aquae rafts, filtered through GF/F glass microfiber filters, and stored at room temperature.
A. flos-aquae rafts were mixed in 100 mL beakers that contained a total volume 70 mL, made up of 68 mL of filtered treatment water and 2 mL of unfiltered UKL lake water containing A. flos-aquae rafts (about 30 rafts per 2 mL). Rafts were transferred to each beaker with a disposable transfer pipette. Each treatment was mixed continuously for 2 h (except for duration of light exposure treatments) with a stir plate, which provided slow vertical and horizontal movement to homogenize the water in each treatment replicate. Beakers were suspended in a water bath at 20 ± 0.1 C. Each beaker was individually illuminated from above with a 20 W 12 V halogen lamp at a distance of approximately 15 cm. Lamp distance was adjusted to achieve 300 μmol photon/m2/s at the water surface (except for the light intensity treatments). Surface water light intensity was calibrated for each lamp with a LI-COR 1400 flat sensor.
Light intensity and mixing time were determined from a preliminary study of the effects of light intensity and mixing on the physical structure of rafts (data not included here). It has been noted that A. flos-aquae rafts from UKL physically change as they separate into individual filaments with too much agitation (Geiger et al. 2005). Our preliminary study found that mixing with a bubbler in small volumes created too much agitation. Slow mixing with a stir plate on the lowest setting did not induce physical changes to the rafts at 2 h, but physical changes were discernible after longer periods of mixing. Our preliminary study also showed that intensities above 300 μmol photon/m2/s for 2 h led to physical changes and drastic buoyancy loss (i.e., many rafts on the bottom of the beaker that could not be tested for buoyancy loss or turgor pressure). At intensities below 300 μmol photon/m2/s for 2 h, turgor pressure differences were also difficult to determine due to a majority of the rafts floating at the water surface. It was therefore determined that 2 h of slow mixing at 300 μmol photon/m2/s provided the best conditions for evaluating buoyancy changes.
After the completion of each treatment, we followed an identical procedure that started with the removal of the beaker from the water bath, immediately followed by a 10 min dark-adapted period. The purpose of the dark-adapted period was to halt active photosynthesis that would result in turgor pressure changes during turgor pressure determination (MacIntyre et al. 1997). After the dark-adapted period, the vertical positions of the rafts in the beaker were recorded using a digital camera in minimal light for determination of initial surface accumulation and depth of suspended rafts (detailed methods to follow). Finally, the contents of each beaker were placed into the pressure chamber for turgor pressure determination.
Surface accumulation and depth of suspended rafts
Surface raft counts were recorded before the first pressure increment application. The number of individual rafts were counted visually after each dark-adapted period and then verified using the digital images. Rafts were considered to be surface rafts if any part of the raft was touching the surface lens of the water. Depths of suspended rafts were determined using the digital video and the tracking software Logger Pro from Vernier Software and Technology. After digitally marking the depth of the center of each suspended raft that was not a surface raft, we calculated the average depth of suspended rafts for each replicate after treatment.
External pressure application and treatment comparison
Cellular turgor pressure was compared by applying external pressure at increments of 0.2 atm using a Bergeon 5555/98, a commercially available pressure chamber. After each pressure increment, A. flos-aquae rafts at the surface, suspended in the water column, and on the bottom of the chamber were counted. Due to the macroscopic size of the A. flos-aquae rafts and the transparent walls of the pressure chamber, counts were completed visually and in real time. Counts were then used to create a buoyancy loss curve of percent buoyant rafts after each applied pressure increment. This method was adapted from Walsby (1971), who used percent collapsed vesicles after each applied pressure increment to create a pressure collapse curve. Pressure collapse curves were used by Walsby (1971) to compare cellular turgor pressure induced from preconditioning of cells. Cells with higher turgor pressure after preconditioning lost vesicles due to collapse at lower applied pressures, so cells with pressure collapse curves that occurred at lower applied pressures had higher turgor pressure. In the study presented here, we compared treatments in the same manner but used buoyancy loss curves instead of pressure collapse curves. To establish that the buoyancy loss curve method would capture the same trend as pressure collapse curve method, we validated the buoyancy loss curve method using manipulations of light intensity and duration of light exposure similar to those performed by Walsby (1971), Grant and Walsby (1977), and Thomas and Walsby (1985). See the section “Light intensity and duration of light exposure treatments” in this paper for a complete description of the adapted method.
Walsby (1971) established the relationship between applied external pressure and turgor pressure, but did not relate changes in turgor pressure to changes in water column position or overall buoyancy of macroscopic rafts. Later work by Walsby showed that cyanobacteria lose buoyancy due to vesicle collapse, but noted that buoyancy loss may not occur until 50% of vesicles have collapsed, so loss is not a 1:1 or linear relationship with vesicle collapse (Thomas and Walsby 1985). For this study, visual counting allowed us to relate turgor pressure increases to water column position (i.e., suspended in the water column or at the surface). After each treatment, the rafts were spread throughout the water column of the treatment beaker (Fig. 2). The large size of the rafts allowed us to count the surface rafts, the rafts suspended in the water, and the rafts on the bottom of the pressure chamber container. During the pressure application, we observed that the last rafts to lose buoyancy in each pressure test were the rafts at the surface. Thus, we compared the turgor pressure of the surface rafts in each treatment by using the pressure that caused the last surface raft in each treatment replicate to lose buoyancy and sink to the bottom of the container. This surface raft buoyancy loss pressure provided a measurement to directly compare the turgor pressure and buoyancy state of surface rafts after each treatment.
Published online:
02 October 2018Figure 2. Profile images of representative replicates from each group after 10 min dark-adapted period. Images show the very top of the water column and cut off just above the bottom of the buoyancy chamber. Grid lines = 0.6 cm.
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/ulrm20/2018/ulrm20.v034.i04/10402381.2018.1466011/20190117/images/medium/ulrm_a_1466011_f0002_c.jpg"})
Figure 2. Profile images of representative replicates from each group after 10 min dark-adapted period. Images show the very top of the water column and cut off just above the bottom of the buoyancy chamber. Grid lines = 0.6 cm.
Light intensity and duration of light exposure treatments
The light intensity and duration of light exposure treatments followed the same protocols and methods as described above for the small-scale laboratory treatments, with the exception of variations in light intensity or light exposure duration (respectively). The light intensity treatments consisted of 6 replicates mixed for 2 h at intensities of 90, 150, 300, and 500 μmol photon/m2/s. The duration of light exposure treatments consisted of 6 replicates mixed under a light intensity of 300 μmol photon/m2/s for 30, 60, 120, and 180 min. After mixing, the A. flos-aquae rafts from each treatment were analyzed using buoyancy loss curves from external pressure application and compared with the results of previous studies that used pressure collapse curves to compare cyanobacteria cells after manipulations of light intensity and duration of light exposure (Walsby 1971, Grant and Walsby 1977, Thomas and Walsby 1985).
Wetland water and ion addition treatments
For the wetland water treatments, rafts of A. flos-aquae were mixed with 3, 10, 33, and 100% wetland water by volume. Treatment mixtures consisted of wetland water and lake water filtered through GF/F glass microfiber filters. The control group consisted of A. flos-aquae rafts mixed with filtered lake water only. Wetland water was collected from the Wood River Wetland (WGS84 42°34′57.0864″N, 121°58′15.4128″W; Fig. 1) each day of the study at sundown, filtered, and stored at room temperature in amber bottles overnight. Replicates of the wetland water treatment levels and the lake water controls were conducted over 2 seasons. In the first season, 24 replicates of each treatment level including the control were performed over 6 d between 27 June and 2 July 2014, with 4 replicates of each treatment completed on each day of the experiments. In the second season, 10 replicates of each treatment level, including the control, were performed over 8 d between 11 July and 21 July 2015, with one replicate of each treatment completed on each day of the experiments.
Treatment mixtures for potassium, sodium, and calcium ions were performed over 8 d between 22 and 30 June 2016 with one replicate of each treatment completed on each day of the experiments. The target ion concentrations of the treatment mixtures were based on the results from 2014 and 2015 wetland water studies, which indicated the most effective wetland dose for controlling buoyancy was the 10% treatment. Concentrations of K+, Na+, and Ca2+ for filtered UKL water and the 10% wetland treatments were determined during the wetland water addition experiments in 2015. Concentration of each ion was determined using a Dionex ICS-5000 ion chromatography system fitted with an Ionpac CS12A analytical column. To prepare the ion treatment replicates, Hanna Instruments 0.1 M certified stock solutions of K+, Na+, and Ca2+ were added to filtered UKL water to reach target concentrations. For each ion, A. flos-aquae rafts were mixed in 3 treatment levels: low-level, mid-level, and high-level for each ion type (). The mid-level concentration was targeted to be equivalent to the 10% wetland concentration for each ion type. To understand the combined effect of these ion types, a fourth combination mixture group was added. The combination mixtures consisted of low-level, mid-level, and high-level treatments from all 3 ions combined together for each treatment level (e.g., the low-level treatment was a combination of the low-level additions for potassium, sodium, and calcium).
Table 1. Concentration of sodium, potassium, and calcium in UKL lake water and low-level, mid-level, and high-level treatments after additions. Units in mg/L.
Data analysis
Data analysis was conducted in R (R Core Team 2014). Data was checked for normality using the Shapiro–Wilk test and each data group was compared to a normal distribution using the Kolmogorov–Smirnov test (ks.test function in R). Equal variance for each group was compared using the var.test function. Data transformation did not improve normality and equal variance for most data sets and groups in this study, so all data was analyzed without transformation. When assumptions for normality and equal variance were not violated, groups were compared by one-way ANOVA using the functions aov and lm. The linear model (lm) function was used to compare all the groups together and the ANOVA function (aov) was used to compare each treatment group with the control group individually. When assumptions for normality and equal variance were violated, groups were compared by the Kruskal–Wallis rank sum test using the kruskal.test function.
Results
Light intensity and duration of exposure treatments
Cellular turgor pressure increased as light intensity or exposure time increased. For the light intensity treatments, buoyancy loss curves and surface raft buoyancy loss (Fig. 3) showed that as light intensity increased, A. flos-aquae rafts lost buoyancy at progressively lower pressures. The buoyancy loss curves for the duration of light exposure treatments and surface raft buoyancy loss (Fig. 3) both indicate that as exposure time increased, A. flos-aquae rafts lost buoyancy at progressively lower externally applied pressures. This trend of buoyancy loss at lower applied external pressures indicates that the turgor pressure in A. flos-aquae cells increased as light intensity and duration of exposure increased.
Published online:
02 October 2018Figure 3. Buoyancy loss in A. flos-aquae rafts after manipulated light treatments. Curves show average percent rafts still floating at each externally applied pressure for light intensity treatments (A) and for duration of light exposure (B). Points show average applied pressure at which the last surface raft lost buoyancy after light intensity treatments (C) and for duration of light exposure (D). Error bars represent 95% confidence interval. Light intensity units: μmol photon/m2/s.
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/ulrm20/2018/ulrm20.v034.i04/10402381.2018.1466011/20190117/images/medium/ulrm_a_1466011_f0003_b.jpg"})
Figure 3. Buoyancy loss in A. flos-aquae rafts after manipulated light treatments. Curves show average percent rafts still floating at each externally applied pressure for light intensity treatments (A) and for duration of light exposure (B). Points show average applied pressure at which the last surface raft lost buoyancy after light intensity treatments (C) and for duration of light exposure (D). Error bars represent 95% confidence interval. Light intensity units: μmol photon/m2/s.
Wetland treatment groups
Cellular turgor pressure in A. flos-aquae rafts was highest in the 10% wetland water treatment group. The surface rafts required less applied pressure to induce sinking compared to all other treatment groups in both study years (Fig. 4) and significantly less (P < 0.05) pressure than the control group in both years (). The average depth of the suspended rafts was also the lowest for the 10% treatment (Fig. 5) and significantly lower (P < 0.05) than the control group in both years (). These 2 pressure measurements indicate that the average cellular turgor pressures of rafts in the 10% treatment groups were higher than the rafts in all other treatment groups and the control group.
Published online:
02 October 2018Figure 4. Buoyancy loss in A. flos-aquae rafts after wetland water treatments. Curves show average percent rafts still floating at each externally applied pressure for the wetland water treatments in 2014 (A) and 2015 (B). Points show average applied pressure at which the last surface raft lost buoyancy after wetland water treatments in 2014 (C) and 2015 (D). Error bars represent 95% confidence interval. Star symbol indicates statistically significant difference from the control group (0% wetland water) at P < 0.05.
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/ulrm20/2018/ulrm20.v034.i04/10402381.2018.1466011/20190117/images/medium/ulrm_a_1466011_f0004_b.jpg"})
Figure 4. Buoyancy loss in A. flos-aquae rafts after wetland water treatments. Curves show average percent rafts still floating at each externally applied pressure for the wetland water treatments in 2014 (A) and 2015 (B). Points show average applied pressure at which the last surface raft lost buoyancy after wetland water treatments in 2014 (C) and 2015 (D). Error bars represent 95% confidence interval. Star symbol indicates statistically significant difference from the control group (0% wetland water) at P < 0.05.
Published online:
02 October 2018Figure 5. Average of percent rafts floating at the surface after the 10 min dark-adapted period for 2014 (A) and 2015 (B) and average depth of rafts suspended in the water column (but not at the surface) for replicates in each group for 2014 (C) and 2015 (D). Error bars represent 95% confidence interval. Star symbol indicates statistically significant difference from the control group (0% wetland water) at P < 0.05.
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/ulrm20/2018/ulrm20.v034.i04/10402381.2018.1466011/20190117/images/medium/ulrm_a_1466011_f0005_b.jpg"})
Figure 5. Average of percent rafts floating at the surface after the 10 min dark-adapted period for 2014 (A) and 2015 (B) and average depth of rafts suspended in the water column (but not at the surface) for replicates in each group for 2014 (C) and 2015 (D). Error bars represent 95% confidence interval. Star symbol indicates statistically significant difference from the control group (0% wetland water) at P < 0.05.
Table 2 Summary of results for the analysis of the average applied external pressure to cause the surface rafts to lose buoyancy after the wetland water treatments. Group means represent the pressure at which the last raft lost buoyancy in each treatment. Experimental groups with added wetland water were compared to the control group (0% wetland water).
Table 3. Summary of results for the analysis of average floating position. Group means represent the average depth of rafts not at the surface. Experimental groups with added wetland water were compared to the control group (0% wetland water).
The cellular turgor pressure results in the rafts of the 3% and 33% treatments were similar, lower than the 10% treatment group, and higher than the control group. The applied pressure required to cause surface rafts to sink in these 2 treatments was nearly identical in both study seasons (). The average suspension depth results were also similar in 2014, but varied in 2015 (). However, the average depths of the rafts in the replicates of both groups were deeper than the control group, but shallower compared to the 10% treatment group. This data indicates that the 3% and 33% treatments induced a similar level of turgor pressure in the cells of the A. flos-aquae rafts and the induced turgor pressure was higher than cells of the rafts in the control group and lower than the 10% treatment.
The cellular turgor pressures of the rafts in the 100% wetland water treatment group were consistently lower than all other treatment groups and the control group in this study. This result is most easily seen on the buoyancy loss curves (Fig. 4) where the 100% treatment group buoyancy loss curve occurs at distinctly higher externally applied pressures over the length of the curve in both study years. This is also evident from the pressure required to cause surface raft buoyancy loss (Fig. 4, ). While the surface raft buoyancy loss pressures were not statistically higher than the control group in either 2014 or 2015 for the 100% treatment, the results do indicate that turgor pressure was reduced by the 100% wetland water treatment. Additionally, the average depths of the suspended rafts were lower than the controls in both years (), which indicates that the turgor pressures were higher than in the control group for the nonsurface rafts of this treatment. It is not clear if the 100% treatment induced a lower turgor pressure for all A. flos-aquae rafts compared to the control group, but the 100% wetland water did induce a lower turgor pressure for surface rafts in these studies because they consistently sank at lower applied pressures than the control group. It is important to note that the turgor pressure induced by the 100% treatment was far lower than all of the other treatment groups (3%, 10%, and 33%).
The comparison of the surface raft accumulation showed the same overall trend as the turgor pressure comparisons for each treatment group (Fig. 5, ). The 10% treatments had a lower percentage of rafts at the surface than all of the other treatments and had a significantly lower (P < 0.05) percentage of surface rafts than the control group in both study years. The 100% wetland treatment group had a higher percentage of surface rafts than all of the other treatment groups in both study years and was significantly higher (P < 0.05) than the control group in 2015, with an average of 51.8%. The 3% and 33% treatments groups had a similar average percentage of surface rafts that were below the average of the 100% group in both years and above the average of the 10% group in both years. In 2014, the 3% and 33% treatment groups had an almost identical average percentage of rafts. In 2015 the percentage of surface rafts for the 33% group was much higher than the 3% treatment groups and showed a higher average than the control group. Overall, the surface comparison shows the 10% wetland water treatment was the most effective at reducing A. flos-aquae surface accumulation, the 100% treatment was the least effective, and the 3% and 33% treatments showed an effect between the 10% and 100% treatments. Images of typical replicates from each treatment (Fig. 2) show low surface accumulation in the 10% treatment, high surface accumulation in the 100% treatment, and an intermediate level of surface accumulation in the 3% and 33% treatments.
Table 4. Summary of results for the percent surface raft analysis. Group means represent the average percent of rafts at the surface. Treatment groups with added wetland water were compared to the control group (0% wetland water).
Ion addition treatments
The ion addition treatments indicated that all treatment levels induced an increase in cellular turgor pressure in A. flos-aquae rafts compared to the control group. Buoyancy loss curves for each ion treatment level occurred at distinctly lower applied pressures than the control group (Fig. 6). The externally applied pressures required to cause surface raft buoyancy loss were lower, on average, than the control group for each ion addition treatment level (Fig. 6) and significantly lower (P < 0.05) for all combination treatment levels (). Additionally, the mean suspended raft depths were significantly lower (P < 0.05) than the control group for all treatment levels except for the mid-level sodium treatment (P < 0.10). All of these results indicate that each ion addition treatment induced an increase in cellular turgor pressure.
Published online:
02 October 2018Figure 6. Buoyancy loss in A. flos-aquae rafts after ion addition treatments. Curves show average percent rafts still floating at each externally applied pressure for the ion addition treatments of sodium (A), potassium (C), calcium (E), and the ion combination (G). Points show average applied pressure at which the last surface raft lost buoyancy after ion addition treatments of sodium (B), potassium (D), calcium (F), and the ion combination (H). Potassium concentration used for combination plot, but each combo treatment consisted of additions of all 3 ions. Error bars represent 95% confidence interval. Star symbol indicates statistically significant difference from the control group (lake water with no additions) at P < 0.05.
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/ulrm20/2018/ulrm20.v034.i04/10402381.2018.1466011/20190117/images/medium/ulrm_a_1466011_f0006_b.jpg"})
Figure 6. Buoyancy loss in A. flos-aquae rafts after ion addition treatments. Curves show average percent rafts still floating at each externally applied pressure for the ion addition treatments of sodium (A), potassium (C), calcium (E), and the ion combination (G). Points show average applied pressure at which the last surface raft lost buoyancy after ion addition treatments of sodium (B), potassium (D), calcium (F), and the ion combination (H). Potassium concentration used for combination plot, but each combo treatment consisted of additions of all 3 ions. Error bars represent 95% confidence interval. Star symbol indicates statistically significant difference from the control group (lake water with no additions) at P < 0.05.
Table 5. Summary of results of ion addition treatments. Treatment groups with ion additions were compared to the control group (no add treatment). ANOVA results only shown for the pressure at which surface raft lost buoyancy (surface raft loss column). Surface raft % and mean raft depth data were analyzed using Kruskal–Wallis rank sum.
The ion concentrations chosen for this study did not have a progressive effect on cellular turgor pressure with successive additions. The buoyancy loss curves for sodium and potassium were almost identical for all treatment levels (Fig. 6), indicating that their effect on turgor pressure was similar for all treatment levels. The calcium buoyancy loss curves were distinct over most pressures with the high-level occurring at the lowest applied pressures. However, the low-level calcium treatment occurred at lower applied pressures than the mid-level treatment, indicating that it induced a higher turgor pressure. Also, all calcium treatment level curves were almost identical over the last section of the curve. This section of the curves corresponded to the surface rafts, indicating that all calcium treatment levels induced a similar level of turgor pressure in the surface rafts. For the ion combination buoyancy loss curves, the low-level treatment occurred at the lowest applied pressures, which indicates that the low-level treatment was most effective at inducing turgor pressure. Overall, the ion treatment levels used in this study produced similar buoyancy loss curves indicating they had similar effects on turgor pressure for each ion treatment type, not an increasing effect with successive additions.
The comparison of percent surface rafts shows that for all treatment levels of each ion addition type showed that there were significantly fewer (P < 0.05) surface rafts than in the control group (). However, similar to the applied external pressure analysis, there was not an inverse effect based upon the treatment level concentration. Each ion addition type showed similar results between each treatment level (Fig. 7). The high-level treatment had the fewest surface rafts for the sodium, potassium, and combination additions, the mid-level treatment had the highest surface raft percent for the potassium and combination treatments, and the high-level treatment had the highest surface raft percent for the calcium treatment; no overall trend based upon treatment level was shown. Each treatment level led to similar surface raft accumulations and was not significantly different than other treatment levels within each ion addition type (statistical comparison data not shown).
Published online:
02 October 2018Figure 7. Average of percent rafts floating at the surface after ion addition treatments of sodium (A), potassium (C), calcium (E), ion combination (G), and average depth of rafts suspended in the water column after ion addition treatments of sodium (B), potassium (D), calcium (F), and ion combination (H). Depths have units of centimeters below surface. Potassium concentration used for combination plot, but each combo treatment consisted of additions of all 3 ions. Error bars represent 95% confidence interval for each group. Star symbol indicates statistically significant difference from the control group (lake water with no additions) at P < 0.05.
![]({"type":"image","src":"/na101/home/literatum/publisher/tandf/journals/content/ulrm20/2018/ulrm20.v034.i04/10402381.2018.1466011/20190117/images/medium/ulrm_a_1466011_f0007_b.jpg"})
Figure 7. Average of percent rafts floating at the surface after ion addition treatments of sodium (A), potassium (C), calcium (E), ion combination (G), and average depth of rafts suspended in the water column after ion addition treatments of sodium (B), potassium (D), calcium (F), and ion combination (H). Depths have units of centimeters below surface. Potassium concentration used for combination plot, but each combo treatment consisted of additions of all 3 ions. Error bars represent 95% confidence interval for each group. Star symbol indicates statistically significant difference from the control group (lake water with no additions) at P < 0.05.
Even though the results of each ion treatment were similar, the most consistently effective ion treatment type was the combination treatment. As a group, the combination treatments resulted in fewer surface rafts, a lower depth of suspended rafts, and loss of surface rafts at lower externally applied pressures when compared to the treatments consisting of only one ion type (sodium, potassium, or calcium). Within the ion combination group, the low-level treatment showed the most consistent overall effect on A. flos-aquae raft buoyancy. The buoyancy loss curve for the low treatment level occurred at a lower applied pressure than any other ion type treatment level. The suspended rafts had a lower average depth (3.01 cm) than all other ion type treatments except for the low-level potassium treatment (3.02 cm). The surface raft percent (2.91%) was lower than all other treatments, except for the high-level combination treatment (2.74%). The applied pressure to cause surface raft buoyancy loss (1.69 atm) was lower than all other treatment levels, except for the high-level combination treatment (1.65 atm). These results all indicate that the ion combination treatments induced consistently higher turgor pressure and surface raft reduction than any single ion type additions, and that among the combination treatment levels, the low-level treatment was the most effective at inducing turgor pressure and reducing surface raft accumulation.
Discussion
Pressure collapse and buoyancy loss method comparison
The results of the light intensity manipulation trials using the buoyancy loss method to measure the effects of external pressure application captured the same overall trend as the pressure collapse curve method developed by Walsby. The initial description of the pressure collapse method compared Dolichospermum flos-aquae cells in “normal” light (7000 lux) with cells in low light (300 lux) conditions, and showed that the gas vesicles in cells in higher light conditions collapsed at lower externally applied pressures than cells in normal light conditions (Walsby 1971). Our manipulations here showed that rafts exposed to higher light conditions sank at lower externally applied pressures than did rafts in lower light conditions (Fig. 3). This result indicates that both studies capture the same overall trend of higher cellular turgor pressure from increased light intensity conditions. While we used a higher number of light intensities, shorter exposure times, and a different buoyant CyanoHAB taxa than Walsby (1971), experiments by Kromkamp et al. (1986) used the pressure collapse method on A. flos-aquae using similar exposure times and intensities to our study and showed the same overall trend of increased cellular turgor pressure as light intensity increased.
The light duration manipulation trials using the buoyancy loss method to measure the effects of external pressure application also captured the same overall trend as the pressure collapse curve method developed by Walsby. The original explanation of the pressure collapse method (Walsby 1971) did not include trials of light duration but later publications using the pressure collapse method did show that a lower externally applied pressure was required to cause vesicle collapse as exposure time increased. Thomas and Walsby (1985) showed this for Microcystis cells exposed to 160 μmol photon/m2/s for times between 1 and 5 h and Kromkamp et al. (1986) showed this for A. flos-aquae filaments exposed to 100 and 200 μmol photon/m2/s for 2 and 5 h. The conclusion of both of these studies was that an increase in the duration of the exposure to light led to an increase in cellular turgor pressure. Our light duration manipulations showed the same overall trend of an increase in cellular turgor pressure as the duration of light exposure increased for A. flos-aquae rafts using the buoyancy loss method (Fig. 3) at 300 μmol photon/m2/s from 30 min to 3 h.
Allison and Walsby (1981) used the pressure collapse curve method to show that extracellular potassium concentration led to an increase in turgor pressure for Dolichospermum flos-aquae. They compared one treatment level of potassium (1 mM K+) with a control group (no K+ in the solution) at 200 μmol photon/m2/s for 145 min. The cells suspended in the 1 mM K+ solution had a distinct curve that occurred at much lower pressures than the control with no crossover from start to finish. Each of the potassium addition treatments in this study showed the same results as Allison and Walsby (1981), a distinct curve at lower applied pressures than the control using the buoyancy loss curve method. While increasing the concentration of the potassium additions in this study did not result in a decrease in pressure to cause buoyancy loss, the method captured the result of an increase in turgor pressure from potassium addition when compared to controls.
The overall goal of this part of our study was to adapt the gas vesicle pressure collapse curve method of Walsby (1971) to a method using a commercially available apparatus and a visual assessment of initial position and sinking for the results of the pressure test (instead of relying on a colorimeter for refractive index analysis). Based upon the results of these experiments and how they relate to the results of Walsby (1971), Allison and Walsby (1981), and Thomas and Walsby (1985), we conclude that this adaptation was successful. The overall trends shown by Walsby for light intensity, light duration, and potassium additions were shown here using the buoyancy loss method, which supports this method adaptation as a success, and validates its use for further experiments. This research also added the comparison of surface raft turgor pressure using the externally applied pressure to cause the last surface raft in each replicate to lose buoyancy. Due to the fact that surface rafts are particularly resistant to buoyancy loss, the comparison of the pressure to induce surface raft buoyancy loss between treatments is meaningful in this study because it provided direct insight into the effect of changing conditions and water treatments on the turgor pressure of cells in the rafts at the water surface.
Wetland treatment groups
The results of this study show that a mixture of wetland water and lake water from UKL containing A. flos-aquae rafts in small-scale treatments led to an increase in cellular turgor pressure, a decrease in raft suspension depth, and a reduction in raft surface accumulation. The most effective treatment level was the 10% wetland water treatment. The hypothesis for the wetland water treatments was for surface accumulation of rafts to decrease and for cellular turgor pressure to increase as the concentration of wetland water increased, showing a maximum effect in the 100% treatment. However, the effect of wetland water on A. flos-aquae turgor pressure and surface accumulation did not show a cumulative effect and the 100% treatment group was the least effective, thus our hypothesis for the wetland water treatments was not supported by the data. Data from treatments during both years of the wetland additions showed that the 100% treatment group had a higher average surface raft accumulation and a consistently lower average cellular turgor pressure than all of the other treatment groups and the control group. The results from both years also showed that the rafts in the 3% and 33% treatment levels were very similar to each other, with values falling consistently between the rafts in the 100% treatment and the rafts in the 10% treatment. These results indicate that the 10% wetland water addition could represent a target treatment level that causes A. flos-aquae rafts to be suspended lower in the water column instead of dominating the light environment at the surface.
The fewer surface rafts, lower depth of suspended rafts, and lower applied pressure to cause buoyancy collapse in the 10% treatments imply that the conditions created by the 10% wetland water treatment led to A. flos-aquae buoyancy optimization adjustments in the water column. Optimization through buoyancy regulation, floating and sinking to adjust water column position, is dependent on existing conditions and influenced by many factors (Walsby 1994). One of those factors is phosphate availability because it has a direct effect on the formation of gas vesicles (Oliver 1994). Konopka et al. (1987) showed that gas vesicle content in phosphate-limited cultures of A. flos-aquae remained low, even when light intensity was high. When phosphorus concentration was increased, vesicle content also increased. Kromkamp et al. (1989) showed the same relationship between phosphorus and gas vesicles for cultures of Microcystis aeruginosa when phosphorus was limiting and when phosphorus concentrations were increased. This finding is important for the results presented here because the wetland water used for this study had a much higher content of soluble reactive phosphorus (SRP) than lake water at the collection site. Measurements of SRP from UKL at the collection site averaged 15 μg/L compared to an average of 197 μg/L for the 10% treatment and 1340 μg/L for the 100% wetland water treatment. We did not measure gas vesicle content or synthesis in this study, but this drastic difference in SRP between the lake water and wetland water indicates that gas vesicle formation could have been triggered by the addition of wetland water. However, the process of triggering and producing gas vesicles takes up to 24 h to take full effect (Walsby 1994), so it is unlikely that this factor influenced buoyancy and water column position of A. flos-aquae rafts in this study due to the 2 h length of each treatment.
Another factor that influences sinking and floating is light attenuation. Gas vesicle construction is repressed when light intensity is high (Thomas and Walsby 1985) and triggered when light intensity is low (Oliver 1994). This detail could be an important factor in our study because wetland water greatly attenuates light due to high concentrations of dissolved organic compounds (Eloranta 1999) that can greatly increase the extinction coefficient and lower overall irradiance (Lean 1998). Based upon this relationship, addition of wetland water to the treatments in this study led to an increase in light attenuation and a decrease in overall irradiance in the treatment beakers. Increased surface accumulation of A. flos-aquae rafts in the 100% wetland treatments of this study could have been influenced by this effect, but similar to the role of SRP in surface raft accumulation, the level of influence is unclear for this study due to the 2 h length of time for each treatment. Triggering and constructing gas vesicles due to light attenuation can take up to 24 h (the same overall mechanism that is triggered by high phosphate), so the length of the treatments here makes it likely that other factors were more influential to the results of this study.
Reduced irradiance also decreases cellular turgor pressure (Walsby 1994). Two main processes that lead to increases in turgor pressure in cells of cyanobacteria, carbohydrate accumulation and ion exchange, are both driven by photosynthesis (Kromkamp et al. 1986, Oliver 1994). Reduction in overall irradiance has been shown to decrease cellular turgor pressure in many species of buoyant cyanobacteria (Thomas and Walsby 1985, Oliver 1994) and specifically in A. flos-aquae (Kromkamp et al. 1986). For this study, the results support that a decrease in cellular turgor pressure due to reduced irradiance could have occurred in the 100% wetland treatments due to absorption by humic substances in the dark brown water. Average cellular turgor pressure of rafts in the 100% treatment groups was lower than all other treatments and the controls for all turgor pressure analyses in 2014 and 2015. The results also show that this did not occur in the 3%, 10%, or 33% treatment groups because the average cellular turgor pressure for all of these groups was higher than controls in both years. Due to the fact that changes in irradiance have been shown to influence cellular turgor pressure in A. flos-aquae in 2 h in other studies (Kromkamp et al. 1986), it is likely that the reduction in irradiance due to additions of wetland water had an effect on turgor pressure in this study. The effect, however, is not definitive since there are other factors that also influenced cellular turgor pressure in this study.
The last factor to discuss here, which could have influenced cellular turgor pressure of A. flos-aquae in this study, is ion exchange. This mechanism is an essential component of cyanobacteria photosynthesis because it facilitates the uptake of carbon and the efflux of hydrogen ions (Kaplan et al. 1989). Cyanobacteria release hydrogen ions from cells during photosynthesis in order to alleviate pH imbalance through a Na+–H+ antiport protein (Blumwald et al. 1984). This exchange is considered to occur in all cyanobacteria taxa, and has been specifically measured for Synechococcus and Dolichospermum (Sherer et al. 1988, Kaplan et al. 1989). Cyanobacteria also bring monovalent ions (i.e., sodium and potassium) into cells through symport proteins along with bicarbonate ions in order to supply carbon for photosynthesis (Kaplan et al. 1984). Both of these processes accumulate ions in cyanobacteria cells during active photosynthesis, leading to an increase in turgor pressure. These are key processes for this study because they can be influenced by the external concentration of sodium and potassium ions, with cellular turgor pressure having been shown to change due to manipulations of the extracellular concentration of these ions (Allison and Walsby 1981, Reed and Walsby 1985). Therefore, ion exchange is a likely influence on the increase in turgor pressure of A. flos-aquae cells in this study, making the results of the individual and combination ion treatments important to consider.
Ion addition treatments
The hypothesis for the ion addition treatments was for surface accumulation of rafts to decrease and for cellular turgor pressure to increase as the concentration of ion additions increased, showing a maximum effect in the high-level addition treatment for each ion type. Similar to the results of Allison and Walsby (1981), extracellular addition of cations led to an increase in cellular turgor pressure. This result occurred for all of the individual ion additions (sodium, potassium, and calcium) and the combination of all 3 ions. However, turgor pressure did not increase progressively as ion concentration increased as was expected. Many of the treatment levels had similar results for all treatment levels, despite an increase in concentration. This lack of a progressive trend is seen best in the pressure collapse curves of potassium and sodium (Fig. 6), which show a clear separation between the curves of each addition group from the control group but not a clear separation between the curves of the individual addition groups. The most distinct separation between treatment level pressure collapse curves occurred in the combination additions for this study, but the separation did not follow the expected trend, either. The low-level addition showed the lowest average sinking pressure instead of the high-level addition. Overall, the low-level ion addition treatments showed similar results to the mid-level and high-level addition treatments, indicating that a threshold may exist at or below the low-level concentration additions used in this study.
The ion addition treatments were included in this study to elucidate the mechanism that led to increased turgor pressure and decreased surface accumulation in the wetland water treatments. The concentrations of sodium, potassium, and calcium in each of the low-level additions in this study were targeted to be halfway between the lake water and the 10% wetland water treatment. As stated in the methods, the mid-level ion addition treatments were targeted to have the same concentrations as the 10% wetland water addition treatments from 2015. Our calculations for the concentration of the additions were based upon concentrations determined from lake water at the collection site in July of 2015. Sample ion concentrations at the collection site were determined to be higher in 2016 than they were in 2015. This difference led to the concentration of the low-level treatments in the ion addition treatments to be more comparable than the mid-level treatment additions to the ion concentrations in the 10% wetland water additions in 2014 and 2015.
The proximity of sodium, potassium, and calcium concentrations between the low-level ion additions and the 10% wetland water mixture is important here because the low-level addition combination treatment and the 10% wetland treatments were the most effective treatments. When the buoyancy collapse curves between the 2015 wetland water treatments (Fig. 4) and the combination ion treatments (Fig. 6) were compared, the low-level combination treatment occurred at lower applied pressures and the mid-level and high-level treatments occurred at slightly higher applied pressures. This trend is similar to the trend of the 10% and 33% treatments for the wetland water additions. This result indicates that the concentration of sodium, potassium, and calcium ions influenced the cellular turgor pressure and surface accumulation results of A. flos-aquae rafts in this study. However, the effect was not analogous to the wetland water treatments for the surface raft accumulation and surface raft turgor pressure. All combination treatments were virtually identical for surface raft percent and the pressure to cause surface raft buoyancy loss. This result was different from the 10% and 33% wetland water treatments, which showed a discernable difference between surface raft accumulation and surface raft turgor pressure (Figs. 4 and 5). Clearly ions play a key role in the effectiveness of the 10% wetland water treatment but there are factors beyond these 3 ions that are leading to an increase in turgor pressure and a reduction of surface raft accumulation.
Application to lake management
The underlying goal of this work was to reduce the surface accumulation of buoyant CyanoHABs in order to diminish their physical dominance at the top of the water column in a small-scale study in order to later apply this method on a lake-wide scale. Wetland water and ion additions used in this study did not completely eliminate surface rafts, but the number of rafts at the surface was reduced and the depth of the remaining suspended rafts was expanded over a larger depth creating less surface density. Computer models of vertical algal movement and distribution show that concentrated surface layers of buoyant cyanobacteria lead to appreciable shading (Klausmeier and Litchman 2001). A. flos-aquae has a very low critical light intensity, so shading from concentrated layers of surface rafts creates conditions that are ideal for growth, allowing A. flos-aquae to outcompete other phytoplankton species to become the dominant taxa (Huisman et al. 1999). When the surface accumulation depth of buoyant cyanobacteria is expanded in computer simulations, the photic zone is expanded, allowing for better competition for light between phytoplankton taxa (Klausmeier and Litchman 2001). Higher incident light that allows for better competition between species reduces the advantage of low-light growth by some buoyant cyanobacteria and often leads to their displacement as the dominant taxa by green algae or other buoyant cyanobacteria (Huisman et al. 1999). Thus, methods to reduce surface accumulations of cyanobacteria could lead to higher algal species richness and reductions of the negative impacts from CyanoHABs on a lake-wide scale.
The practical use of this strategy would be to apply wetland water or ion additions to a lake in order to obviate CyanoHABs surface blooms. The use of this addition as a management strategy would be as a preventative measure, before a bloom occurs, and not as an active treatment method for the reduction of a surface bloom after it forms. This strategy would most likely not completely prevent CyanoHABs in lakes with consistently heavy blooms, but could reduce their effects or delay their onset. For lakes such as UKL, with consistently heavy blooms of A. flos-aquae, surface reduction would allow for greater phytoplankton diversity, which could reduce the unhealthy water quality effects associated with thick surface blooms that can lead to lethally low oxygen events and native fish die-offs. A delay in bloom formation could also reduce or prevent the onset of microcystin-producing blooms of Microcystis aeruginosa, which form after early-season and mid-season blooms of A. flos-aquae elevate nitrogen levels in the system (Eldridge et al. 2017). Large-scale manipulations have been shown to lead to a delay in the onset of A. flos-aquae blooms in other systems, delaying or preventing late-season Microcystis blooms (McDonald and Lehman 2013), so a large-scale ion addition or wetland water addition could produce a comparable, desirable outcome in UKL.
Conclusion
The addition of wetland water and cations in this study increased cellular turgor pressure in A. flos-aquae rafts, but the treatment levels did not increase turgor pressure or decrease surface accumulation as expected. For the wetland water, there was a maximum effect at an intermediate treatment level and for the cation additions all treatment levels had similar results, indicating a threshold effect. While these results were unexpected, they do indicate that mixing wetland water with UKL water and the addition of cations to UKL water does disrupt A. flos-aquae buoyancy. Furthermore, the lower than expected effective treatments shown here would be more practical for lake-wide management than the expected high-level treatments. As a small-scale study, the results may not yet apply to a lake-wide management. However, these results provide a promising foundation for further study on a larger scale that could ultimately be used as a lake-wide management strategy. Due to its history of wetlands and its shallow depth, UKL is an optimal location for this strategy. However, this approach could be applied to any system with adjacent wetlands with an adequate volume to reach the target dose. These research findings could be a first step toward a management strategy that could include actions that reduce surface accumulation of buoyant CyanoHABs through induction of turgor pressure. Ideally, this strategy would be used along with current management practices of nutrient reduction and nutrient ratio manipulation in order to achieve the best results for reducing surface accumulation and inhibiting CyanoHAB dominance.
| Lake | Low-level | Mid-level | High-level | |
|---|---|---|---|---|
| Sodium | 9.46 | 12.46 | 15.46 | 18.46 |
| Potassium | 2.06 | 2.62 | 3.22 | 3.82 |
| Calcium | 6.88 | 7.73 | 8.58 | 9.43 |
| 2014 trials | 2015 trials | |||||||
|---|---|---|---|---|---|---|---|---|
| Wetland water (%) | n | Mean pressure (atm) | Kruskal–Wallis rank sum results | n | Mean pressure (atm) | Kruskal–Wallis rank sum results | ||
| χ2 | P | χ2 | P | |||||
| 0 | 24 | 2.28 | NA | NA | 8 | 2.30 | NA | NA |
| 3 | 24 | 2.15 | 1.945 | 0.1631 | 8 | 2.14 | 3.6018 | 0.0577 |
| 10 | 24 | 2.09* | 5.9766 | 0.0145 | 8 | 1.92* | 9.3299 | 0.0023 |
| 33 | 24 | 2.13 | 3.34 | 0.0641 | 8 | 2.08* | 5.5395 | 0.0186 |
| 100 | 24 | 2.43 | 2.1536 | 0.1422 | 8 | 2.40 | 1.336 | 0.2477 |
* Indicates statistically significant difference from the control group (0% wetland water) at P < 0.05.
| 2014 trials | 2015 trials | |||||||
|---|---|---|---|---|---|---|---|---|
| Wetland water (%) | n | Depth avg. (cm) | ANOVA results | n | Depth avg. (cm) | ANOVA results | ||
| F | P | F | P | |||||
| 0 | 24 | 2.43 | NA | NA | 8 | 2.35 | NA | NA |
| 3 | 24 | 2.66* | 6.193 | 0.0165 | 8 | 2.53 | 0.583 | 0.458 |
| 10 | 24 | 2.83* | 19.76 | 0.00006 | 8 | 2.81* | 7.679 | 0.015 |
| 33 | 24 | 2.69* | 8.883 | 0.0046 | 8 | 2.67 | 2.738 | 0.12 |
| 100 | 24 | 2.49 | 0.531 | 0.47 | 8 | 2.55 | 0.675 | 0.425 |
* Indicates statistically significant difference from the control group (0% wetland water) at P < 0.05.
| 2014 trials | 2015 trials | |||||||
|---|---|---|---|---|---|---|---|---|
| Wetland water (%) | n | Surface rafts (%) | ANOVA results | n | Surface rafts (%) | ANOVA results | ||
| F | P | F | P | |||||
| 0 | 23 | 23.6 | NA | NA | 12 | 34.6 | NA | NA |
| 3 | 22 | 21.0 | 0.809 | 0.373 | 8 | 24.8 | 1.368 | 0.257 |
| 10 | 22 | 17.5* | 4.53 | 0.0391 | 8 | 18.9* | 5.719 | 0.0279 |
| 33 | 23 | 21.9 | 0.364 | 0.549 | 8 | 35.2 | 0.075 | 0.787 |
| 100 | 21 | 27.8 | 0.923 | 0.342 | 8 | 51.8* | 5.624 | 0.029 |
* Indicates statistically significant difference from the control group (0% wetland water) at P < 0.05.
| ANOVA results surface raft loss | ||||||
|---|---|---|---|---|---|---|
| Ion addition treatment | n | Surface raft (%) | Mean raft depth (cm) | Surface raft loss (atm) | F | P |
| no.add | 8 | 16.11 | 2.31 | 2.00 | NA | NA |
| na.low | 8 | 6.64** | 2.78** | 1.850 | 1.2353 | 02851 |
| na.mid | 7 | 5.70** | 2.58* | 1.714* | 3.3333 | 0.0910 |
| na.high | 8 | 4.94** | 2.96** | 1.800 | 1.6471 | 0.2202 |
| k.low | 8 | 4.31** | 3.02** | 1.850 | 0.797 | 0.387 |
| k.mid | 8 | 7.41** | 2.81** | 1.700* | 3.15 | 0.0977 |
| k.high | 8 | 3.98** | 2.76** | 1.725* | 3.544 | 0.0807 |
| ca.low | 7 | 6.32** | 2.75** | 1.971 | 0.028 | 0.869 |
| ca.mid | 7 | 9.15** | 2.67** | 1.829 | 1.023 | 0.33 |
| ca.high | 7 | 9.18** | 2.73** | 1.743 | 1.623 | 0.225 |
| combo.low | 7 | 2.91** | 3.01** | 1.686 ** | 5.519 | 0.0353 |
| combo.mid | 8 | 3.08** | 2.74** | 1.700 ** | 4.846 | 0.045 |
| combo.high | 8 | 2.74** | 2.78** | 1.650 ** | 6.236 | 0.0256 |
* Indicates statistical significance at P < 0.10. ** Indicates statistically significant difference from the control group (0% wetland water) at P < 0.05.
Acknowledgments
Support was provided for this study by the Center for Lakes and Reservoirs at Portland State University, The Nature Conservancy in Klamath Falls, and the Bureau of Land Management in Klamath Falls.