**4. Discussion**

The databases used in this systematic review were Scopus, Science Direct, and Scielo, as mentioned above. Scopus was chosen because it is widely known as one of the largest databases of abstracts and citations of peer-reviewed literature and has many records in the science area. In addition, it is easy to export bibliographic information for further analysis [25]. Likewise, ScienceDirect is a database with an extensive record of article records in various areas of science [26]. In the case of Scopus, this database focuses more on article records of researchers from South American countries [27]. With the choice of these three databases, the aim was to address the largest number of research studies on picocyanobacteria worldwide, since this is a subject that has not been studied extensively, as has been recognized by several authors.

In this regard, and after conducting the search strategy, it was found that the ScienceDirect and Scopus databases provided the largest number of publications due to their worldwide positioning as indexed databases and their mainly Englishlanguage journals. In addition, Scopus covers various areas of science, technology, medicine, social sciences, arts, and humanities [25]. Moreover, Sciencedirect is a database that also covers multidisciplinary scientific areas [28], however, it is limited to journals and books published directly by its publisher [29]. Therefore, although many of the articles included in this review were found, they did not exceed those found in Scopus [30].

In contrast, the Scielo open-access database, although it includes journals from all areas of science, only two articles associated with the topic were found in the search. This is since this database contains scientific articles published only in Latin America, and because it is a database that publishes mainly in Spanish and Portuguese [31].

There are different molecular techniques used for the identification of picocyanobacteria, in this review we found that the application of these techniques to characterize and amplify portions of the cyanobacterial genome has increased considerably in recent years. These techniques have proven to be valuable for comparing the structures of complex microbial communities, inferring phylogenetic relationships, and monitoring their dynamics in relation to environmental factors [32]. Cyanobacteria such as *Synechococcus* and *Cyanoothece* are particularly difficult to identify and classify [33], most molecular methods to identify them are based on total DNA or RNA extraction and amplification by PCR as shown in **Table 3**. However, there are biases related to the presence of PCR inhibitors and primer specificity and efficiency that can skew the results of community composition [33].

Concerning the number of annual publications obtained in the analyzed period (**Table 3**), there was consistency with the findings of Rousso et al. [4] in their research on predictive models for cyanobacterial blooms in freshwater lakes. They found that in the period between 2014 and 2019 the highest number of publications on cyanobacteria was reported, the same as this research, where it was found that between 2015 and 2019 the highest number of publications on picocyanobacteria was collected. However, the maximum number of publications found for the articles found that met the inclusion criteria was only 5 for the year 2019, indicating that there are still not many studies on picocyanobacteria [34]. Research on cyanobacteria appears to be strongly related to advances in monitoring technology, i.e., increased availability of data, knowledge of cyanobacterial ecology, physiology, and risks, among other factors [34]. Furthermore, due to the environmental problems associated with cyanobacteria and their potentially toxic blooms. Merel et al. [8] evidence that articles on cyanobacteria have increased significantly in the period 1995–2010, a trend that is expected to continue [8]. This systematic review shows that 2010 was a year where no significant reports on picocyanobacteria were found, which probably indicates that research is still focused on microplankton cyanobacteria instead of picocyanobacteria.

Regarding the countries where the studies were conducted, it was found that the United States and China have been outstanding countries for the number of scientific publications on cyanobacteria and their toxic blooms, this is demonstrated by the study conducted by Bertone [4], which analyzed the publications on CyanoHAB in different countries and found that most of the publications are focused on the United States, Northern Europe, Southeast China, Japan, and Oceania [4]. This result coincides with that found in this study (**Figure 2**), where the highest number of scientific articles on picocyanobacteria have been published in the United States and China.

The concentration of publications in developed countries such as these may be related to their economy and extensive scientific resources, the provision of funds for research and development, and the availability of data used for these purposes [35]. Also, these countries have managed to develop specialized monitoring and control procedures for CyanoHAB from research [8, 36]. On the other hand, Ndlela [37], made an overview of cyanobacterial bloom occurrences and research in Africa during the last decade and found that the amount of information available on the continent on the subject is limited probably due to the general inadequacy of the infrastructure and its relation to civil wars [37].

Regarding the most frequent genera, the genus *Synechococcus* was the most reported with a frequency of 24; this genus plays a fundamental role in the ecology of surface water bodies that are important human resources, being predominant in freshwater systems. Generally, picocyanobacteria of the genus *Synechococcus*, *Prochlorococcus,* and *Cyanobium* are designated as non-flourishing [38]. However, some strains of the genus *Synechococcus* can produce toxins such as β-N-methylamino-L-alanine (BMAA), and microcystin (MC) [39], which causes problems in the ecosystem and human health. Similarly, Gin [15] through his study showed that *Synechococcus* spp. could produce cylindrospermopsin (CYN) and anatoxin-a (ATX) which are alkaloids that can cause damage to mammalian organisms, this discovery has implications on the potential risk to freshwater resources that serve as drinking water supply [40].

Previous studies by Li [21] report that the prevalence of *Synechococcus* in water bodies is influenced by warm temperature, high nutrient level, and phosphorus limitation, comprising fractions of up to 80% of the total biomass of picocyanobacteria of a bloom [41]. Furthermore, the result obtained in this review agrees with that reported by Cabello [38], where it is confirmed that the genera *Synechococcus* and *Cyanobium* are the dominant picocyanobacteria in freshwater systems [42].

*Prochlorococcus* ranks third as one of the most frequently found picocyanobacteria in research. It inhabits the entire photic zone and can be found as deep as 200 m below the surface, being abundant in oligotrophic systems [43]. *Prochlorococcus* and *Synechococcus* can coexist in water bodies, but *Synechococcus* tolerates a wider temperature range, without being limited by temperatures as low as 2°C and is more ubiquitous and has a wider latitudinal distribution [32].

It has been shown that the trophic state of water bodies influences the composition and abundance of picocyanobacterial communities. It was observed that the most studied lakes were those in an oligotrophic state, these lakes are characterized by being poor in nutrients and having low primary productivity [44], which limits the presence of a high microbial density and only those taxa that have adapted to these conditions can survive. Thus, picocyanobacteria of the genus *Synechococcus* are predominant in these systems, this is due to the ability of these picocyanobacteria to adapt to low light conditions, their affinity for orthophosphate and other sources of inorganic phosphorus, as well as their ability to store nitrogen in phycobilins that increase the competition of *Synechococcus* against algae and other bacteria, as stated by Vanstein [45]. The above is consistent with that reported by Joachim Ruber et al. [46], who describe this important genus as dominant in oligotrophic conditions, concluding that *Synechococcus* could be used as a bioindicator in such environments [46]. Besides, authors such as Rousso et al. [4] reported in a systematic review on CyanoHAB that more than 50% of the lakes investigated were eutrophic or hypertrophic and only 8% of the lakes were oligotrophic, reporting that the occurrence of CyanoHAB is related to the levels of nutrients present in the lakes [47].

In **Figure 5**, it is evident that the distance between two keywords demonstrates relative strength and similarity of topic and circles in the same color group suggest that a similar topic is being addressed among the publications [48]. **Figure 5** shows that the most used keywords are: *Synechococcus*, microbial community, phylogeny, and 16S rRNA, this shows us that more molecular identification strategies have been used recently for the identification of picocyanobacteria as cited by Demoulin et al. [49], who indicate that since the late 1990's many phylogenetic studies based on 16S rRNA or specific protein have been published. Similarly, the results of the keywords are also observed in **Figure 5**, in which the most frequent words have been used a greater number of times in the articles. The total link strength attribute indicates when a

keyword is very important because it is identified to have had a lot of interaction with other keywords in the analyzed articles, the higher the value the stronger the link that exists between one word and another [50].

The findings of the current systematic review show the lack of research on picocyanobacteria in surface waters that allow understanding the importance they represent as beneficial microorganisms; standing out for being part of the primary producers, or harmful because they can produce toxic blooms. It was also evidenced that molecular identification methods of picocyanobacteria have recently begun to be highlighted in research methodologies, which shows a transition from traditional research to a more advanced one.

### **5. Conclusions**

Although in the last two decades the identification of picocyanobacteria has increased due to the implementation of new automated methods and molecular techniques, studies aimed at identifying them in surface water bodies intended for recreational use or drinking water supply are still incipient, which is possibly explained by the difficulty in their characterization and rapid physiological plasticity. The predominant genus of picocyanobacteria in this systematic review was *Synechococcus*, a producer of toxic compounds, which generates an alert and highlights the importance of advancing in the implementation of protocols for sampling and identification of these bacteria for epidemiological surveillance.

The countries where more studies on cyanobacteria were conducted were the United States and China since these are developed countries that invest their resources in education and research and can develop specialized monitoring and control procedures for CyanoHAB from their scientific resources. Therefore, there is a need for further research in this area, to use the information for further studies and decision making.

### **Acknowledgements**

The authors would like to thank the support provided by the Research Group on Health and Sustainability of the University of Antioquia, as well as the School of Microbiology who, through the financing of projects for degree work, made possible the realization and completion of this project.
