The Complicated and Confusing Ecology of Microcystis Blooms

Blooms of the toxin-producing cyanobacterium Microcystis are increasing globally, leading to the loss of ecosystem services, threats to human health, as well as the deaths of pets and husbandry animals. While nutrient availability is a well-known driver of algal biomass, the factors controlling “who” is present in fresh waters are more complicated. Microcystis possesses multiple strategies to adapt to temperature, light, changes in nutrient chemistry, herbivory, and parasitism that provide a selective advantage over its competitors.

still unable to answer the common question raised by the citizen constituents: "Why do we get Microcystis blooms?" Here, we present a discussion on the many factors influencing bloom formation, persistence, and decline and the research efforts required to understand them.
Bottom-up controls. For decades, monitoring efforts have focused on nutrient concentrations as a predictor of phytoplankton biomass. Yet static nutrient concentrations in lakes are as much the residual of biological transformations as they are a cause of blooms. Moreover, all algae need N and P to support carbon fixation, although not all algae assimilate all chemical forms of N and P with similar efficiencies. For example, it is well-known that most marine Prochlorococcus do not possess nitrate reductase genes allowing for the assimilation of nitrate. A freshwater parallel may be the assimilation of urea, which has been used increasingly in recent decades as an agricultural fertilizer (3,11). Urea is an effective N source for many organisms, including Microcystis (12,13). Indeed, the ability to use urea as an N source has been touted as one of the advantages that Microcystis has over competing plankton.
Microcystis bloom events commonly increase surface water pH to well above 9 as the cyanobacterium rapidly consumes available inorganic carbon (14). Under these conditions, the availability of dissolved CO 2 to phototrophs is negligible, and even bicarbonate concentrations are low. Numerous researchers have noted that cyanobacterial carbonic anhydrase gives Microcystis an advantage in the use of bicarbonate as a carbon source; it should be noted that cyanobacteria are also well adapted to high CO 2 concentrations (14). However, recent work has demonstrated that urea can also serve as a carbon source for Microcystis at alkaline pH (15), offering another selective advantage. Moreover, at pH conditions Ͼ9.26, ammonium is converted to ammonia that can diffuse from the system in gaseous form, making more stable N species (e.g., urea and nitrate) important and decreasing total water column N. Perhaps even more importantly, the success of Microcystis in raising the pH can create conditions unfavorable for other phytoplankton, e.g., the siliceous frustules of diatoms become soluble, and Si is likely incorporated at lower rates under these pH conditions. That said, pH swings due to rampant photosynthesis are diel processes, yielding shifts of up to 0.5 pH units (15). Typically, pH decreases at night due to respiration without coincident CO 2 uptake, and thus, diatom success may be linked to whether frustule synthesis can occur at night. Given the paucity of diel studies on bloom gene expression (16,17), whether pH alone can lead to the exclusion of diatoms requires further investigation. Perspective ® We note this competition is not restricted to diatoms; pH, along with nutrients and temperature, is a potential driver that can promote Microcystis success (or lack of) over other algal taxa and bloom-forming cyanobacteria such as Planktothrix, Dolichospermum, and Cylindrospermopsis (18). Little is known about the factors that constrain their interactions and the outcomes, although predation, as well as nutrient and anthropogenic loads, likely play key roles (1,19) These observations highlight a key point with respect to Microcystis populations: how they compete with one type of organism (e.g., diatoms) is likely different from how they compete with another (e.g., other cyanobacteria).
Increasing temperatures provide another condition that favors some cyanobacteria. Microcystis populations grow faster at warmer temperatures (20). Yet toxin production by Microcystis cells responds opposite to this trend; Microcystis strains in culture produce less toxin per cell when grown at warmer temperatures, consistent with field observations, where blooms accumulate less toxins as the season progresses (21). While reduced toxin production has been linked to the loss of microcystin-producing genotypes from populations (22), the mechanisms that could drive a seasonal and specific gene loss (or selection for populations) remain unclear (especially when that gene returns in subsequent years). Other factors, including the possible role of microcystins in offsetting oxidative stress in cells and the effects of lower temperatures increasing

FIG 2
The seasonal cycle of a cyanobacterial bloom in a large dimictic lake. The availability of nutrients (N, P, and Si), dissolved CO 2 , and pH conditions are suggested by the position of acronyms above (high nutrient concentrations or high pH) or below (low nutrients and dissolved CO 2 , lower pH) black marker lines. Acronym positions are relative (no scale implied). Bloom formation in many lakes starts as temperatures increase and stores of nutrients from the winter begin to be consumed and are depleted. As nutrients are depleted and blooms form, cyanobacteria like Microcystis are able to drive down CO 2 concentrations using nutrients that may not be accessible to other planktonic phototrophs. This reduces available CO 2 and increases pH. As temperatures decrease in fall months, dimictic lakes turn over and "reset" the system. Perspective ® excitation pressure/photoinhibition and the production of oxygen radicals, seem more plausible given the high cellular Fe quota driving Fenton chemistry and the presence of photosensitized pigments (23). Regardless of the mechanism, observations point to the confounding issues of growth, temperature, and toxin production; at lower temperatures (ϳ18°C), populations have slower growth rates, producing lower biomass, yet cells produce more toxins. At warmer temperatures (ϳ25°C), biomass is higher, but the toxin cellular quota decreases. Thus, a seasonal shift in toxicity occurs as temperatures increase into the late summer months. Yet temperatures across seasons are not linear; daily swings of 1 to 3°C in the surface mixed layer are common (15), and increased episodic weather associated with climate change (24) may cause water temperature fluctuations that could lead to bursts of toxin production (21). These confounding variables also point to a scientific conundrum: to protect public interests, ecosystem stewards must focus on the toxin per volume water (concentration), as that is where causative issues lie. Yet for scientists to elucidate why Microcystis makes toxins, they need to be focusing on cell quotas to understand the process.
Removal processes. Accumulation of Microcystis biomass also depends on removal mechanisms, namely, grazing, parasites, and virus-mediated lysis. Lab studies have demonstrated that Microcystis cells are selectively rejected as pseudofeces by filterfeeding mussels in a process that indirectly promotes Microcystis growth (25). Moreover, despite Microcystis-specific phage occurring at densities that reach 10 5 ml Ϫ1 , these cyanobacteria proliferate at high cell densities for extended periods (26). Part of their secret may be in the establishment of a lysogenic relationship with phage; in some other prokaryotes, lysogeny imparts homoimmunity to infections by related viruses (27). Yet episodes of viral lysis have been suggested to release intracellular microcystins into the dissolved phase, complicating water treatment protocols (28). The incoming toxin load can be reduced by flocculation of bloom biomass at the water plant intake, whereas dissolved microcystins bypass this step and require more costly chemical treatment(s) (23). Understanding patterns of lytic versus lysogenic infection and factors contributing to lysogen induction will be useful in developing best practices for water utilities. Moreover, research focusing on the extent to which pathogens influence bloom composition and toxicity is required to enable predictive models. Recently, viruses have been shown to play another understudied role: viral infection of competing plankton may provide Microcystis with an advantage. In a slight reinterpretation of the "kill-the-winner" model (29), the presence of viruses infecting competing plankton (e.g., diatoms) may provide another selective advantage for Microcystis populations (30).
Microcystis has numerous advantages over competing plankton in lakes and can both tolerate and exploit conditions of pH, nutrient availability, temperature, and predation that constrain other plankton. Such conditions depend on season and location, so ecosystem managers and researchers must recognize that each factor may contribute to Microcystis success in different ways at different sites around the world. Beyond the N versus P debate regarding constraints on ecosystem productivity (3), research must also focus on what competing plankton cannot do or tolerate in working to understand why Microcystis has become globally successful. Moving forward, a balance between laboratory work with cyanobacterial isolates, mesocosm manipulations, and fieldwork examining microbial community dynamics will be critical as the effects of competition with other algae, constraints imparted by the co-occurring microbial community (phycosphere), and shifting pressures due to climate change are addressed. Yet for all these complications, integration of data from molecular biology and physiology with remote-sensing of increasingly "smart" lakes (31) will provide a path forward in the protection of our most valuable natural resource: clean, potable water.