Microbial communities living in freshwater ecosystems (as well as brackish and marine) are closely related to water quality and are key drivers of aquatic biogeochemistry. Some freshwater microorganisms are potential pathogens, while others may produce toxins^1,2,3. In turn, freshwater microbial communities are influenced by nutrients, pH, temperature, salinity, light intensity, etc. as it is affected by environmental conditions as well as human activities.⁴. Despite the importance of freshwater ecosystems, they remain relatively under-sampled in time and space^5,6,7,8. To ask how aquatic microbial communities differ between different water bodies, we sampled 57 different freshwater bodies in Israel (Figure 1). These water bodies were selected to cover several different characteristics of water bodies with different water uses and distributions and were sampled at different time scales, resulting in three different datasets (Figure 2). Dataset 1 has not been presented elsewhere, datasets 2 and 3 contain metagenomic data for samples previously analyzed in the context of community structure (16 S rRNA gene amplicon sequencing^9,10).

Location, satellite images (from govmap.gov.il) and photographs of sampling sites. The color coding of the fields corresponds to the rainfall gradient in (mm).

Schematic layout of the three datasets. Color coding (similar to Fig. 1 (mm) corresponds to the precipitation gradient.

Dataset 1: Various freshwater bodies with different characteristics

The first set of data is from various bodies of water collected in late summer or early fall when water levels are relatively low (November 2014).^{11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28}. The following sites were sampled (Figure 1): (i) Dalton (DAL^{11,12,15,16,17,18,19,20}), an irrigation reservoir that receives water from several sources, including rainwater, saline springs, and recycled wastewater. These waters are used for irrigation and the area has a history of algal blooms that can clog irrigation systems; (ii) Ein Afek (EA^13,14), water nature reserve. Due to the intensive use of natural springs, the reserve’s water level is controlled, for example, by purchasing water from the national network; (iii) Reshafim Reservoir (Resh^21,22), a multi-purpose reservoir that mainly receives spring water and is used for both irrigation and aquaculture; (iv) Timorim Sol (Tim_L^25,26), primary sewage pool, (v) Timorim right (Tim_R^23,24), tertiary treated sewage reservoir (the water is then used for agriculture); (vi) Yerucham (ER^27,28), a desert reservoir that receives both rain and treated wastewater from the watershed (Fig. 1). These sites were selected to encompass many different features and use ponds of freshwater bodies, some of which have a history of toxic cyanobacterial blooms.^29,30,31,32. From one of these sites, DAL, we also sequenced metagenomes from March, May and November of the following year to provide a seasonal period (2015). In each sampling campaign, water was collected from the top 15 cm and sequentially filtered through 5 μm and 0.22 μm filters representing particle-associated and free-living bacteria, respectively (more details below). In total, this database includes 18 metagenomes (Table 1).

Dataset 2: two-year time series in a highly impacted reservoir

The Dor Aquaculture Research Unit is a research facility used for both aquaculture research and intensive research^{33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55}semi-commercial growth^9,56,57. We monitored microbial populations in several ponds at monthly intervals over three years at this site and showed that seasonality is a major driver of planktonic microbial populations in this system, despite being highly influenced by aquaculture activities.⁹. Here we present metagenomes sequentially from a 23-month sample over two consecutive years (2013–2014, Table 2). These samples were collected from the top 15 cm and filtered on a GF/F filter (nominal pore size 0.7 μm). A full description of the study site and environmental conditions, as well as analysis of 16 S rRNA gene amplicons from the same samples, is provided by Marmen. and b.⁹.

Dataset 3: Winter bloom in Lake Kinneret dominated by Microcystis sp

Lake Kinneret, located in northern Israel, has been well studied for decades^{58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73}including the dynamics and causes of winter Microcystosis to flourish^74,75. We collected 16 samples at four time points spanning a dynamic period in terms of environmental conditions and dominant phytoplankton layers during the winter and spring of 2015 (Table 3). For a detailed description of these samples as well as analysis of 16 S rRNA gene amplicons, see ref.¹⁰. This dataset contains different sampling depths: (i) depth-integrated samples from four dates collected with a 20 μm vertical grid (0–15 m);^58,59,66,67; (ii) Discrete depth samples from two dates (0 (surface)^64,65,72,731^62,63,70,71 and 5 meters^60,61,68,69two dimensional fraction as above).

Metadata

These metagenomic datasets are accompanied by various environmental measurements. The methods are briefly presented here, more details are available in the information section⁹. Briefly, electrical conductivity (EC), dissolved oxygen, temperature and pH were measured. in place using field probes (Eutech instruments, Singapore). Ammonia, total phosphorus, NO₂NO₃ and PO₄ measured with an AA3 Segmented Flow Multichemistry Analyzer (SEAL Analytical, Germany) according to the manufacturer’s protocols (Ammonia: Method No. G-327-05 Rev. 7 – Fluorescent method; Nitrate and Nitrite: Method No. G-172 -96 Rev. 17 and Phosphate: Method No. G-297-03 Rev. 5). Toxin concentrations were measured using the Microcystins/Nodularins (ADDA) Elisa from Abraxis according to the manufacturer’s protocol.

Photosynthetic pigments were quantified using ultra-performance liquid chromatography (UPLC) with a method adapted to the LOV protocol.⁷⁶. Briefly, pigments were extracted in absolute methanol for 3 h in the dark, filtered through 0.2 µm PTFE membranes (Pall Life Sciences, New York, NY, USA) and heated to 30 °C. 10 µl was injected into an ACQUITY UPLC system (Waters Corporation, Milford, MA, USA) equipped with a photodiode array detector. Separation was performed on a C8 column (ACQUITY UPLC BEH, 50 mm column length) using a linear gradient (solvent A – 70:30 methanol: 0.5 M ammonium acetate; Solvent B – 100% methanol). Peaks were identified based on their retention time and absorption spectra and quantified by comparison with Chlorophyll a, Chlorophyll b, Chlorophyll c standards.₂Zeaxanthin, β-carotene, Diatoxanthin, Dinoxanthin, Fucoxanthin and Peridinin (DHI Laboratory, Hørsholm Denmark).

Extracted DNA and archival samples are available for further research by the community (please contact the corresponding authors). The complete database contains ~416.5 gigabases of raw sequence data (Supplementary Table 2). Datasets 1, 2, and 3 provide opportunities to compare and contrast differences within and among these different freshwater sites in temporal, seasonal, spatial, water quality, and environmental conditions.