The volatilome reveals microcystin concentration, microbial composition, and oxidative stress in a critical Oregon freshwater lake

Lindsay Collart; Duo Jiang; Kimberly H Halsey

doi:10.1128/msystems.00379-23

. 2023 Aug 17;8(5):e00379-23. doi: 10.1128/msystems.00379-23

The volatilome reveals microcystin concentration, microbial composition, and oxidative stress in a critical Oregon freshwater lake

Lindsay Collart ¹, Duo Jiang ², Kimberly H Halsey ^1,^✉

Editor: Rachel Poretsky³

PMCID: PMC10654074 PMID: 37589463

ABSTRACT

Toxins produced by cyanobacterial blooms in freshwater lakes are a serious public health problem. The conditions leading to toxin production are unpredictable, thereby requiring expensive sampling and monitoring programs globally. We explored the potential of volatile organic compounds (VOCs) to indicate microcystin presence and concentration and microbial community composition in Upper Klamath Lake, Oregon. Elastic net regularization regression selected 24 of 229 detected m/z + 1 values (corresponding to unique VOCs) in models predicting microcystin contamination. These models outperformed regression models based only on environmental parameters, including chlorophyll, pH, and temperature. Several m/z + 1 values selected by elastic net were putatively identified as saturated fatty aldehydes, which defend cyanobacteria against oxidative stress. Elastic net also identified unique sets of m/z + 1 values that predicted the relative abundance of the dominant bacterial phyla, classes, and cyanobacterial genera. VOCs appear to reveal the physiological status of cyanobacteria during toxic blooms and may be a key component of lake monitoring strategies.

IMPORTANCE

Harmful algal blooms are among the most significant threats to drinking water safety. Blooms dominated by cyanobacteria can produce potentially harmful toxins and, despite intensive research, toxin production remains unpredictable. We measured gaseous molecules in Upper Klamath Lake, Oregon, over 2 years and used them to predict the presence and concentration of the cyanotoxin, microcystin, and microbial community composition. Subsets of gaseous compounds were identified that are associated with microcystin production during oxidative stress, pointing to ecosystem-level interactions leading to microcystin contamination. Our approach shows potential for gaseous molecules to be harnessed in monitoring critical waterways.

KEYWORDS: cyanobacteria, volatile organic compounds, microcystin, harmful algal blooms, cyanotoxins, saturated fatty aldehydes, HAB monitoring

INTRODUCTION

Cyanobacterial harmful algal blooms (cyanoHABs) occur globally and are characterized by excessive growth of photosynthetic bacteria in freshwater lakes and rivers. CyanoHABs degrade water quality, negatively impacting potability, aquatic life, and agricultural and recreational activities. Some cyanoHABs produce toxins that pose direct threats to animal and human health (1 – 5) and also stunt crop development (6, 7). The chemical ecology of cyanotoxins is not well understood, but they appear to provide grazing defense and protection from reactive oxygen species (ROS) that can stimulate cyanobacterial growth and production (8). Cyanotoxins can also alter the microbial community and disrupt multitrophic interactions (9). Annual economic losses caused by cyanoHABs in the USA alone are conservatively valued at $2–$4 billion (10, 11), and the severity and consequences of cyanoHABs are predicted to be exacerbated by climate change (12 – 16). These widely ranging impacts call for near to real-time monitoring of cyanobacteria and their toxins to protect the public and effectively manage cyanoHABs in source and recreational waters (17).

CyanoHAB monitoring programs are challenged because the specific toxins produced are strain-specific, and no morphological shifts or commonly measured environmental triggers are known to be reliably associated with toxin production (17). Quantifying the genes encoding cyanotoxins in an ecosystem offers one approach to assess the risk of cyanotoxin contamination, but may be of limited value because the presence of cyanotoxin genes is not evidence of its expression (18). Direct cyanotoxin measurement in water and fish tissues by enzyme-linked immunosorbent assay (ELISA) and liquid chromatography-tandem quadrupole mass spectrometry (LC-MS/MS) (19) is expensive, specific to a subset of congeners (20) and cannot identify the producer or its abundance (21). Tools to leverage high-resolution detection of cyanobacterial cells and their metabolites remain nascent in application but are needed to address many environmental problems that are reaching ecological tipping points despite decades of intensive scientific effort (22).

We investigated the potential of the volatilome to provide high-sensitivity detection of cyanoHABs and cyanotoxin production (23 – 25). The “volatilome” is the full range of low-molecular-weight (~30–272 a.m.u.) volatile organic compounds (VOCs) produced in an ecosystem (26). VOCs have roles in cell signaling (27 – 30), predator-prey interactions (31), microbial carbon cycling (32, 33), and atmospheric emissions that impact tropospheric ozone and climate (34). Some VOCs inhibit growth and induce lysis in algal community members (35 – 37), thus regulating microbial interactions and community composition (37 – 39). VOC production depends on the algal species present, their growth phase, and their environment (29, 32, 40 – 42).

Algae, including cyanobacteria, release a wide array of VOCs, including terpenes, fatty acids and their 2-keto acid degradation products, alkanes and alcohols (38, 43 – 45) as a result of primary and secondary metabolism (46, 47) and indirectly through photochemical reactions with dissolved organic matter (48). Cyanobacteria produce medium-chain aliphatic hydrocarbons that strengthen and add flexibility and fluidity to lipid bilayers (49, 50), as well as provide tolerance to temperature and light stress (51, 52). Alkanes (C_n) and their corresponding alcohols and aldehydes are produced from C_{n + 1} fatty aldehydes via aldehyde deformylating oxygenase (ADO) yielding aliphatic hydrocarbons of varying lengths (51). The array of volatile hydrocarbons present in cyanobacteria appears to provide mechanisms for managing cell-level oxidative stress (51).

Upper Klamath Lake (UKL) is a large shallow lake in southern Oregon that is a hub of complex water use for agriculture, wildlife, fisheries, recreation, and Tribal subsistence and culture. Intensive farming and drought have decreased water quality in UKL over the last half century, contributing to annual cyanoHAB events. UKL cyanoHABs are typically dominated by Aphanizomenon and Microcystis and produce the hepatotoxin, microcystin, at elevated concentrations that prompt public warnings to avoid water contact in the mid- to late-summer (53). We characterized the volatilome using proton transfer reaction time-of-flight mass spectrometry (PTR-ToF-MS) at lake and canal sites in UKL over 2 years and identified over 200 m/z + 1 values, corresponding to unique VOCs. Elastic net regularized regression selected small subsets of the m/z + 1 values that were effective predictors of microcystin contamination or microbial community composition in UKL. Microcystin prediction by elastic net models outperformed other models based only on environmental variables and in-water properties that are commonly used to detect cyanoHAB development. Several m/z + 1 values recurring in our elastic net models appear to be associated with the fatty aldehyde ADO pathway, suggesting these cyanobacterial metabolites underlie lipid repair and ROS reduction during oxidative stress, which is thought to be associated with microcystin production. These fatty aldehydes in combination with other key VOCs may indicate ecosystem interactions associated with microcystin production and represent important targets for cyanoHAB monitoring.

RESULTS AND DISCUSSION

Upper Klamath Lake chemical and microbial composition

Water samples were collected from three sites on UKL, one site on its northern arm, Agency Lake, and four canal sites during the months of May–December in 2018 and 2019 (Fig. 1). The mean microcystin concentration among UKL samples with detectable toxin was 8.7 ppb (Table 1), surpassing the U.S. Environmental Protection Agency’s recommended health advisory limit for drinking water of 0.3 ppb for pre-school-aged children and 1.6 ppb for children and adults (54). The minimum reporting, recreational, and drinking water limits for microcystin vary by state depending on water use and potential for exposure (55). Of the 70 samples collected over 2018–2019, 10 UKL samples and three canal samples were contaminated with microcystin at concentrations ≥0.3 ppb. Toxic samples mostly occurred in summer months (July to Sept), but occasionally in November 2019, and occurred at all four lake sites (NAL, WBR, EPP, PEL) (Fig. 2). The highest microcystin concentration was 469 ppb from NAL in September 2019. Environmental parameters varied widely in UKL (Fig. 2; Table 1), and toxic samples were sometimes associated with high temperature, chloride, pH, particulate organic carbon (POC), particulate organic nitrogen (PON), chlorophyll, ammonium (AMM) and conductivity; however, no significant correlations were observed with microcystin concentration and any single parameter measured at UKL (Fig. S1).

Fig 1 — Sample sites on Upper Klamath and Agency Lakes, Oregon. Lake sites and their geospatial positions were NAL (42.559839–121.929579), WBR (42.314529–121.942224), EPP (42.430715–121.962764), and PEL (42.2390–121.8097). Canal sites and their geospatial positions were NCA (42.1222–121.8289), ADY (42.0808–121.8456), MER (42.0536–121.6006), and LRI (42.1733–121.6175).

TABLE 1.

Environmental parameters collected at Upper Klamath Lake, Oregon ^a

Environmental parameter	Abbreviation	Minimum	Maximum	Mean	SD
Microcystin (ppb)	TOX	0	469.51	8.72	56.03
Chlorophyll (µg/mL)	CH	0	6.8	0.24	0.94
Temperature (°C)	TEMP	−0.1	26.87	17.62	6.65
pH	PH	7.22	10.22	8.59	0.94
Conductivity (S/m)	COND	61.4	133.6	102.9	12.6
Particulate organic carbon (µg/mL)	POC	0.45	432.32	18.38	65.53
Particulate organic nitrogen (µg/mL)	PON	0.066	100.20	3.96	15.12
Chloride (ppm)	CHL	2.392	50	4.39	6.94
Sulfate (ppm)	None	1.70	50	4.46	6.97
Nitrate (ppm)	None	0.2	15.54	0.73	2.26
Phosphate (ppm)	None	0.11	50	1.37	7.35
Ammonium (ppm)	AMM	0.01	5.43	0.41	0.82

Open in a new tab

^{^a}

Microcystin concentration is the first listed environmental parameter. Below microcystin, “low-cost” parameters are italicized and “high-cost” parameters are in normal font.

Fig 2 — Unsupervised hierarchical clustering of *m/z +* 1 values selected by elastic net in M1, M2, M7, and M8 (left panel) and in models predicting cyanobacterial genera relative abundances (right panel) in lake and canal samples. Samples are shown in rows and labeled with site and date. The heatmap shows the Pearson correlation between each *m/z +* 1 value and microcystin concentration or relative abundances of cyanobacteria genera, with yellow being the most positively correlated and dark blue being the most negatively correlated (legend: VOC magnitude). To the left of the heatmap are environmental parameters identified by elastic net or stepwise linear regression or logistical models associated with each sample: TOXT, microcystin concentration ≥0.3 ppb (red) or <0.3 ppb (black); TOXC, microcystin toxin concentration (ppb); TEMP, temperature (°C); CH, chlorophyll concentration (mg/mL); pH; POC, particulate organic carbon (μg/mL); PON, particulate organic nitrogen (μg/mL); AMM, ammonium (ppm); and SUL, sulfate (ppm).

Untargeted volatilomics detected 229 m/z + 1 values in samples collected at UKL and associated canals during 2018 and 2019. Six m/z + 1 values were present in significantly discriminating amounts between samples with microcystin ≤0.3 ppb and samples with microcystin ≥0.3 ppb (Fig. 3). Using these six m/z + 1 values in a multiple linear regression model failed to predict microcystin contamination or concentration (R ² = 0.08; P-value = 0.89). Volatilomes clustered well by sampling date, and samples collected in 2018 mostly clustered separately from those collected in 2019 (Fig. S2). Volatilomes of toxic samples did not demonstrate clear clustering (Fig. S2).

Fig 3 — Volcano plot showing log₂ fold changes (FC) and P-values (significance determined via Wald test) for the 229 *m/z +* 1 values in toxic (≥0.3 ppb) versus non-toxic samples. Multiple test correction using the Benjamin-Hochberg false discovery rate was applied to the P-value for each *m/z +* 1 value. The points are colored according to log₂ fold changes and degree of significance. Points with positive fold changes greater than the dashed vertical line at +1.0 are *m/z +* 1 values enriched in toxic versus non-toxic samples. Points with negative fold changes less than the dashed vertical line at −1.0 are *m/z +* 1 values depleted in toxic versus non-toxic samples. NS, *m/z* + 1 value that is not present in significantly different amounts between toxic and non-toxic samples.

The relative abundances of four phyla, Cyanobacteria, Bacteroidota, Pseudomonadota, and Actinobacteria, represented 79%–99% of the 16S rRNA sequences in all UKL samples during 2018–2019 (Fig. 4). Members of the class Cyanophyceae were only ~10% of the microbial community in May and peaked in September 2019 when they were up to 75% of the community before decreasing in the autumn months. The four bloom-forming and potentially microcystin-producing cyanobacteria genera in ULK were Aphanizomenon, Anabaena/Dolichospermum, Microcystis, and Gloeotrichia. Anabaena/Dolichospermum sequences were always the dominant Cyanobacteria, contributing 75%–>99% of the sequences in all samples. The relative abundance of Microcystis represented 5%–25% of sequences in August through December and May, but Microcystis were absent in June and July.

Fig 4 — Coarse description of microbial community composition in UKL samples across 2018–2019. Microcystin concentration is shown as points at the top, with concentrations ≥0.3 ppb in red. Relative abundances of the four toxin-producing cyanobacteria genera (top bar graph), four most abundant microbial classes (middle bar graph), and four most abundant microbial phyla (bottom bar graph). Samples are ordered by date on the x-axis.

Cell morphological characteristics of Aphanizomenon, which is the dominant Cyanobacteria during the mid-summer in UKL (56), were commonly observed in UKL samples inspected by light microscopy (Fig. S3). Nevertheless, few sequences were placed within Aphanizomenon and instead sequences often grouped with representatives of Anabaena sp. strain 90, Dolichospermum circinate strain ACBU02, and Anabaena sp. strain WA 102. 16S rRNA-based phylogenies are so far unable to resolve Aphanizomenon and Anabaena/Dolichospermum (57). For example, the addition of metagenomic data (with morphological validation) from 16 Aphanizomenon, Anabaena, and Dolichospermum strains collected in the Pacific Northwest of the USA to collections of cyanobacterial genomes used in previous phylogenetic analyses still placed some strains, such as an Anabaena strain collected from Washington state, within Aphanizomenon (58).

Microcystin toxin prediction using the volatilome

Elastic net is a regression method that uses regularization to select the input variables that are important for the prediction. We developed elastic net regularized regression models using the UKL volatilome with outputs that were either linearly predictive of microcystin concentration (linear models) or predictive of microcystin concentration ≥0.3 ppb (logistic models) to facilitate different water management approaches (Table 2). Linear model M1 and logistic model M2 were developed using only the 229 m/z + 1 values. Linear model M7 and logistic model M8 were developed using the 229 m/z + 1 values and “low-cost” environmental parameters (e.g., buoy data such as temperature, pH, and conductivity, which are rapidly retrieved by current technologies) (Table 1). Across the four elastic net models, variable selection identified 24 of the 229 unique m/z + 1 as being important to predicting microcystin contamination (Table 3), and their relative concentrations are shown in Fig. 2. Nine m/z + 1 values were selected in two elastic net models, and four m/z + 1 values (151.119, 157.157, 199.189, and 203.185) were selected in three elastic net models (Table 3).

TABLE 2.

Models developed for the prediction of microcystin contamination

Model number	Model type	Input variables	Output type
M1	Linear elastic net	VOCs	Continuous
M2	Logistic elastic net	VOCs	Binary
M3	Linear regression	Low-cost environmental parameters	Continuous
M4	Logistic regression	Low-cost environmental parameters	Binary
M5	Linear regression	Low + high-cost environmental parameters	Continuous
M6	Logistic regression	Low + high-cost environmental parameters	Binary
M7	Linear elastic net	VOCs + low-cost environmental parameters	Continuous
M8	Logistic elastic net	VOCs + low-cost environmental parameters	Binary

Open in a new tab

TABLE 3.

m/z + 1 values identified in models predicting microcystin contamination ^a

m/z + 1 value	Peak variance	Chemical	Shift (m/z)	M1	M2	M7	M8
35.042	0.009346	(CH₄O)H^{+ a}	0.004				x
80.045	0.02135	(C₅H₅N)H^{+ ac} (C₃H₈S)H^{+ a} (C₆H₅)H^{+ a} (C₄H₃N₂)H^{+ b}	−0.004 0.004 −0.008 0.008	C+		C−
83.055	0.022152	(C₄H₆N₂)H^{+ ac} (C₄H₄N₂)H^{+ a} (C₅H₆O)H^{+ ac} (C₃H₄N₃)H^{+ b}	−0.005 0.004 0.006 0.007		C+
83.078	0.022158	(C₆H₈)H^{+ a} (C₆H₁₀)H^{+ ac} (C₅H₈N)H^{+ b}	0.001 −0.008 0.005		C+
98.040	0.026149	(C₅H₄FN)H^{+ c} (C₄H₄N₂O)H^{+ a} (C₃H₃N₃O)H^{+ ac} (C₅H₅O₂)H^{+ b} (C₄H₅N₂O)H^{+ b}	0.000 −0.003 0.005 0.003 −0.008			C+
103.070	0.027491	(C₅H₁₀O₂)H^{+ ac} (C₅H₈O₂)H^{+ a} (C₃H₈N₃O)H^{+ b} (C₄H₈NO₂)H^{+ b}	−0.005 0.006 −0.005 0.007		C+
111.102	0.029633	(C₇H₁₁N)H^{+ a} (C₈H₁₂)H^{+ a} (C₇H₁₂N)H^{+ b}	0.002 −0.006 −0.003			C−	C+
137.129	0.036575	(C₁₀H₁₆)H^{+ acd} (C₁₀H₁₄)H^{+ a}	−0.003 0.005			C−
138.131	0.036842	(C₉H₁₅N)H^{+ ac} (C₁₀H₁₆)H^{+ a}	0.003 −0.005			C+
148.073	0.039494	(C₇H₁₂CIN)H^{+ a} (C₉H₉NO)H^{+ c} (C₉H₇NO)H^{+ a} (C₅H₁₀N₂O₃)H^{+ a} (C₆H₁₁O₄)H^{+ b} (C₇H₉N₄)H^{+ b} (C₂H₇N₆O₂)H^{+ b} (C₆H₁₂N₂Cl)H^{+ b} (C₅H₁₂N₂OP)H^{+ b} (C₄H₁₁N₄S)H^{+ b} (C₅H₁₁N₂OS)H^{+ b} (C₇H₁₂OCl)H^{+ b}	0.003 −0.003 0.006 −0.007 −0.001 −0.002 0.002 −0.004 −0.004 −0.005 0.006 0.008	C+		C−
149.117	0.039772	(C₇H₁₆O₃)H^{+ c} (C₆H₁₄N₂O₂)H^{+ a} (C₈H₁₈S)H^{+ a} (C₁₀H₁₃N)H^{+ a} (C₅H₁₄N₃O₂)H^{+ b} (C₁₀H₁₄N)H^{+ b}	0.000 0.000 0.001 0.002 0.001 −0.003	C−		C+
151.119	0.040306	(C₁₀H₁₄O)H^{+ c} (C₉H₁₄N₂)H^{+ c} (C₉H₁₂N₂)H^{+ a} (C₈H₁₂N₃)H^{+ b}	0.007 −0.004 0.005 0.008	C+	C−	C+
153.095	0.040833	(C₉H₁₂O₂)H^{+ c} (C₄H₁₄N₃OP)H^{+ a} (C₇H₁₀N₃O)H^{+ b} (C₆H₁₀N₅)H^{+ b}	0.004 −0.003 0.005 −0.006			C−
157.157	0.041917	(C₁₀H₂₀O)H^{+ c} (C₉H₁₈N₂)H^{+ a} (C₈H₁₈N₃)H^{+ b}	−0.002 −0.004 −0.001	C−		C−	C−
169.113	0.045106	(C₁₀H₁₆S)H^{+ ac} (C₉H₁₆NO₂)H^{+ b} (C₅H₁₀N₇)H^{+ b}	0.008 0.003 0.005			C−
171.171	0.045655	(C₁₁H₂₂O)H^{+ c} (C₁₀H₂₀N₂)H^{+ a} (C₉H₂₀N₃)H^{+ b}	−0.003 −0.006 −0.003	C+		C−
175.159	0.046718	(C₁₀H₂₂S)H^{+ c} (C₉H₂₀NO₂)H^{+ b} (C₅H₁₆N₇)H^{+ b}	0.008 0.002 0.004	C−		C−
185.185	0.049392	(C₁₂H₂₄O)H^+d (C₁₀H₂₂N₃)H^{+ b} (C₁₁H₂₂NO)H^{+ b}	−0.004 −0.004 0.007	C−		C+
189.151	0.05045	(C₁₄H₁₈)H^{+ a} (C₁₃H₁₈N)H^{+ b} (C₈H₁₈N₃O₂)H^{+ b} (C₁₀H₂₂NS)H^{+ b} (C₄H₁₄N₉)H^{+ b} (C₇H₁₈N₅O)H^{+ b}	−0.004 −0.001 0.003 −0.004 0.006 −0.008			C−
193.153	0.051517	(C₁₁H₁₇N₃)H^{+ a} (C₁₃H₁₈O)H^{+ a} (C₆H₁₈N₅O₂)H^{+ b} (C₁₁H₁₈N₃)H^{+ b} (C₁₂H₁₈N_O)H^{+ b} (C₆H₁₉N₅P)H^{+ b}	0.000 0.003 −0.001 −0.005 0.006 0.007	C−		C−
199.189	0.053128	(C₁₃H₂₆O)H^{+ e} (C₁₂H₂₄NO)H^{+ b}	N/A −0.005		C+	C+	C+
203.185	0.054193	(C₇H₂₀N₇)H^{+ b} (C₁₁H₂₄NO₂)H^{+ b} (C₁₁H₂₅NP)H^{+ b} (C₁₅H₂₂)H^{+ c}	−0.001 −0.004 0.005 0.006	C+	C+	C+
233.959	0.062401	Many compounds			C+
332.868	0.088782	Many compounds			C+		C+

Open in a new tab

^{^a}

C+” indicates that the m/z + 1 value was retained in the model with a positive coefficient, and “C−” indicates a negative coefficient. Shift is the difference between the chemical’s actual mass and detected mass. Chemical identifications were made using the Ionicon PTR viewer integrated database (“a” superscript), the PTR viewer calculated formulas (“b” superscript), GLOVOC database (“c” superscript), previously published PTR-MS research (“d” superscript), or relationships to other identified m/z + 1 values (“e”). Italicized m/z + 1 values were also important in predicting cyanobacterial relative abundance (Fig. 6).

Four additional regression models based on the “low-cost” environmental parameters (M3, M4) or the full collection of environmental parameters (“low + high cost”, M5, M6) were developed to compare against the performance of the VOC-based elastic net models (Table 2). Similar to previous studies (59), “low-cost” linear M3 was weakly predictive of microcystin concentration (MSPE, 19.1) and retained only pH and chlorophyll (Table S1). POC, PON, and AMM strongly boosted the predictive power of linear M5 (MSPE < 1). Neither logistic “low-cost” M4 nor “low + high cost” M6 were able to discriminate whether samples contained microcystin ≥0.3 ppb with greater than 50% accuracy (Fig. 5; Table S2; Fig. S4).

All of the VOC-based models outperformed “low-cost” comparator models to predict microcystin in UKL (Fig. 5). Addition of “low-cost” environmental parameters to the training data did not improve VOC-based model performance (Fig. 5), and except for “month” in M8, were not retained in the final equations (Tables S1 and S2). The high Akaike Information Criterion (AIC) in logistic M2 and M8 are partly attributable to the relatively high number of selected variables (8 and 18, respectively) and were strongly balanced by area under the receiver operating characteristic curve (AUC) values that were 0.78 and 0.88 compared to 0.50 (no better than chance) for M4 and only 0.22 for M6 (Fig. 5; Fig. S4).

VOCs were effective predictors of microcystin in UKL. Our ability to rapidly measure volatile metabolites in water samples requiring no pre-processing (5-minute PTR-MS measurement of raw water samples) provides a unique platform to explore relationships between the volatilome and ecosystem health and the potential for VOCs to be leveraged in cyanotoxin monitoring. Low volatility of toxins, including microcystin, makes their direct detection by PTR-MS unfeasible. Direct toxin measurement by ELISA or mass spectrometry is the current gold standard for monitoring but can, at times, become too expensive for frequent and widespread application across sensitive waterways dependent on timely public health advisories (19). The metabolome is increasingly used to evaluate human health (60 – 62) and ecosystem status, such as shifts in soil microbial ecology (63). Similarly, the success of the volatilome to provide information about microcystin presence and concentration suggests that unique collections of VOCs in UKL are produced depending on cell physiology and community composition.

Predicting microbial community composition using the volatilome

Elastic net models were also developed using the relative abundances of the four most abundant phyla, classes, and toxin-producing cyanobacteria genera as dependent variables and the 229 m/z + 1 values as independent variables. The 12 resulting models selected a total of 71 m/z + 1 values (Table S3). All 12 elastic net models performed well, yielding mean squared prediction errors (MSPEs) that were 0.75–1.02 and SDs that were 0.08–0.54 (Fig. S5). The m/z + 1 value 205.204 was an important predictor of the relative abundance of Cyanobacteria phylum, Cyanophyceae class, and all four Cyanobacteria genera (Fig. 6). Eleven of the 18 m/z + 1 values predictive of the Cyanobacteria phylum relative abundance were also predictive of Cyanophyceae relative abundance and 14 were predictive of the relative abundance of at least one of the Cyanobacteria genera. Similarly, seven of the eight m/z + 1 values predictive of Actinobacteriota relative abundance were predictive of Actinobacteria (class) relative abundance (Fig. 6). Six m/z + 1 values identified in models predicting microcystin concentration were also identified in models predicting the relative abundances of Cyanobacteria genera (Table 3).

Fig 6 — Pearson correlations between the relative abundances of the four toxin-producing cyanobacterial genera (top), four most abundant microbial classes (middle), or four most abundant microbial phyla (bottom) in UKL samples and the associated *m/z +* 1 values identified in elastic net models (shown at top). Outlined boxes are *m/z +* 1 values identified in the model predicting the relative abundance of the taxonomic group in each row. Boxes with a solid outline indicate *m/z +* 1 values in models predicting relative abundances of cyanobacteria genera, class, or phylum. Note that some *m/z +* 1 values predictive of cyanobacteria genera relative abundances are also predictive of Cyanophyceae and Cyanobateria relative abundances. Boxes with a dotted outline indicate *m/z +* 1 values in models predicting the relative abundance of other taxonomic classes or phyla. Pearson’s r value of 1 (green) indicates a positive correlation, and a value of −1 (purple) indicates a negative correlation.

Elastic net regularized regression yielded a collection of VOC-based models that were highly effective at predicting the relative abundance of key cyanobacteria, including Microcystis, which is thought to be the primary source of microcystin in UKL. The success of these models is likely a consequence of seasonal changes in the microbial community composition and taxonomic and physiological differences leading to the collection of VOCs released (32, 64 – 66). We do not know if the VOCs identified here would also be detected in cultures of the different cyanotoxin-producing cyanobacteria. Fundamental differences in metabolism between strains in culture collections could result in unique volatilomes, and the absence of VOCs observed in this study in cultures may be a consequence of in situ conditions rather than a VOC-strain association. Furthermore, the complex interactions between cyanobacteria and ecosystem processes leading to toxin production remain enigmatic and challenging to replicate in the laboratory. Nevertheless, the subsets of VOCs identified using elastic net revealed m/z + 1 values that were maintained through coarse and finer-grained taxonomic groups. These results indicate systematic relationships between volatilomes and microbial taxa in UKL. We are unaware of studies that have leveraged the metabolome to describe microbial community composition; however, neural networks and linear regression approaches are being used to integrate metabolomic, metagenomic, and taxonomic data (67 – 70). In our study, elastic net regularized regression applied to volatilomes yielded models that were strongly predictive of the cyanotoxin and microbial community composition.

Selected m/z + 1 values in our models suggest that those compounds mediate interactions between cyanobacteria, microcystin, and the environment. For example, a sesquiterpene, m/z + 1 203.185, was retained with positive coefficients by three models predicting microcystin and in models predicting relative abundances of Phylum Cyanobacteria, Cyanophyceae, and Anabaena. Sesquiterpene synthases are present in Anabaena species (71), and the recurrence of m/z + 1 203.185 in our models is consistent with the abundance of Anabaena in UKL and the release of sesquiterpenes and microcystin during cyanobacterial senescence (72).

β-ionone was assigned to m/z + 1 193.153 based on its known PTR-ToF-MS target mass (73). m/z + 1 193.153 was retained with negative coefficients in M1 and M7 predicting microcystin and three models predicting the relative abundance of non-cyanobacterial taxonomic groups. m/z + 1 193.153 was positively correlated with phylum Cyanobacteria, Cyanophyceae, and Anabaena (Fig. 6). β-ionone and other norcarotenoids are products of carotenoid oxidation in various cyanobacteria during photo-oxidative stress, including Anabaena, Aphanizomenon, and Microcystis, and inhibit photosystem II in Microcystis (38, 74 – 77). Oxidative stress in UKL may have induced the production of β-ionone in cyanobacteria (78, 79), thereby decreasing Microcystis abundance and microcystin production. Non-toxic Microcystis strains employ peroxidases in response to oxidative stress, but toxic Microcystis strains may produce microcystin to combat mild, chronic oxidative stress (80). The different pathways employed by cyanobacteria to tolerate oxidative stress point to β-ionone as a potentially important compound that mediates interactions within the cyanobacterial community, including microcystin production. β-ionone is also a taste-odor compound in potable freshwater sources (74) that can be rapidly identified using our approach.

m/z + 1 137.129 is likely limonene with the molecular formula (C₁₀H₁₄)H⁺. Limonene is a monoterpene produced by planktonic and benthic cyanobacteria (81). Other compounds with the same m/z + 1 value reported in PTR-MS databases include pinene and linalool, but neither of these terpenes are produced by wild-type cyanobacteria (82, 83). m/z + 1 137.129 was retained with a negative coefficient in M7 and a positive coefficient in the model predicting the relative abundance of Aphanizomenon. m/z + 1 137.129 was also negatively correlated with Microcystis and Gloeotrichia (Fig. 6). Limonene can inhibit photosynthesis (37, 84) and lyse Microcystis aeruginosa (85), suggesting that limonene produced by Aphanizomenon was associated with lower Microcystis abundance and perhaps consequently, lower microcystin concentrations.

The m/z + 1 values 157.157, 171.171, and 185.185 were selected in M1, M7, and M8, respectively, and differ by 14.014 mass units, suggesting these VOCs are products of sequential demethylation activity. A fourth, m/z + 1, 199.189, is 14.004 mass units greater than 185.185 and was retained in M2, M7, and M8 with positive coefficients. The lowest m/z + 1 value in this series, 157.157, was retained with negative coefficients. Chemical formulas for these m/z + 1 values include C10–C13 saturated fatty aldehydes (SFAs), decanal, undecanal, dodecanal (86), and tridecanal (Table 3). Tridecanal is a key marker for Cyanophyceae (87) but is not yet present in PTR-MS chemical databases and has not been reported in PTR-MS-based research. Nevertheless, the longer chained SFAs (C12 and possibly C13) appear to be upregulated in concert with microcystin production.

The associations between SFAs and microcystin concentration in our elastic net models indicate that the relative abundances of SFAs shift during oxidative stress. SFAs accumulate between the lipid bilayers of cyanobacterial thylakoid and cytoplasmic membranes (50, 88) where they contribute to membrane structure and help fine-tune localization of photosynthetic machinery (50) during temperature and light stress (49, 50, 89). Cyanobacteria use an acyl-acyl-carrier-protein (ACP) reductase/ADO pathway to produce fatty aldehydes of decreasing chain length (C_n, C_{n − 1}, C_{n − 2}, … ; Fig. 7). Fatty aldehydes are substrates for aldehyde dehydrogenase (ALDH) yielding fatty acids that can be used to repair membrane lipids damaged by ROS (i.e., hydrogen peroxide, H₂O₂) produced during photosynthesis. ADO, like other diiron oxygenases, appears to be a powerful oxidizing enzyme with a wide substrate range (51). The alkane products of NADH-dependent ADO activity on fatty aldehydes can also serve as electron donors to reduce ROS. This latter reaction is primed by the generation of a Fe^IV-Fe^IV diiron center in ADO by H₂O₂ and alkane oxidation returns the diiron center to the Fe^III-Fe^III state (90). However, alkane-dependent H₂O₂ reduction could also lead to OH^• accumulation causing a deleterious cycle of cell damage. During high light stress, the fatty acid and alkane metabolites of the ADO pathway would be rapidly depleted requiring larger pools of longer chain-length fatty acids to maintain ongoing lipid repair and H₂O₂ destruction. The genes encoding ADO and ALDH were upregulated in the model cyanobacterium, Synechocystis sp. PCC6803 during high light and oxidative stress (91, 92). As the cell’s capacity to repair systems damaged by ROS becomes overwhelmed, longer chained SFAs may accumulate, making them useful targets for microcystin detection by PTR-MS.

Fig 7 — The role of the saturated fatty aldehyde (SFA) oxidation pathway in membrane lipid repair and depletion of reactive oxygen species. SFAs putatively identified by elastic net models predicting microcystin concentration are in yellow (*m/z* + 1 values 157.157, 171.171, and 185.185). SFAs produced by aldehyde-deformylating oxygenase (ADO) are metabolized by aldehyde dehydrogenase (ALDH) producing saturated fatty acids. Blue arrows and arrowheads represent saturated fatty acids used to repair thylakoid (green) and plasma membrane (gray) lipids damaged by ROS (red), represented here by H₂O₂. Photosynthesis (Ps) unavoidably produces ROS. Medium chained alkanes (C10–C12) reduce ROS via ADO activity (see text). Microcystin can protect against ROS (green arrows) by binding to the photosynthetic subunits and by binding to and promoting the production of ROS-degrading enzymes (RDE). Desaturases A-D (DesA-D) unlink glycerol from the fatty acyl moieties in the membrane. The fatty acyl is removed from the membrane by lipolytic enzyme, LipA. Acyl-ACP synthase (Aas) can reattach the fatty acyl to ADO to re-enter the SFA oxidation pathway, represented by a dotted arrow.

One current prevailing hypothesis for a biological role for microcystin in cyanobacterial cells posits that it protects photosystems and peroxidases against oxidative damage (80, 93 – 95). Increased cyanobacterial dependence on the ADO pathway during oxidative stress is consistent with the mechanistic view that ROS can rapidly accumulate to algicidal concentrations even in the presence of microcystin. Although the m/z + 1 values retained in our microcystin models and assigned here to SFAs need to be independently verified using standards or other mass spectrometry approaches (e.g., gas chromatography-mass spectrometry), selection of this collection of related m/z + 1 values in multiple elastic net models predicting microcystin concentration suggests SFAs’ ecological and biochemical interactions with microcystin production (Fig. 7).

The use of VOCs to evaluate microcystin and microbial composition in UKL is time-efficient and could be streamlined or even automated to inform agencies and water managers within a day. The volatilome in water samples was collected directly by PTR-TOF-MS without the need for pre-processing or sorption onto resins. Our models were designed to determine total microcystin concentration and cannot at this time evaluate toxicity, which would require knowledge of the abundances of specific microcystin congeners (96, 97). Evaluation of the volatilome is a holistic and indirect measurement of the ecosystem. Many of the m/z + 1 values identified in our elastic net models provide valuable targets for future study of their roles in cell- to ecosystem-level processes.

Conclusion

The increasing frequency and severity of toxic cyanoHABs in waterways makes new, cost-effective monitoring strategies an urgent task. The ideal monitoring approach would yield information about cyanotoxin identity and concentration, cyanobacterial abundances, and ecosystem health. The VOCs produced in Upper Klamath Lake, OR, provided information about the integrated growth environment and were leveraged using regularized regression to determine microcystin concentration and microbial community composition in UKL water samples. Specific VOCs, including SFAs, may be the smoking gun needed to quickly detect toxin production in freshwater lakes.

Cyanotoxins can now be detected in many waterways that were thought to be pristine, suggesting that the combination of ongoing human activities and climate change is shifting many waterways toward ecological tipping points where HABs and cyanotoxin contamination are reliable annual events. Application of volatilomes and complex data analysis shows their potential for the guidance of water treatment for taste-odor compounds in drinking water, monitoring of toxic and non-toxic cyanoHABs, and novel discovery of ecological interactions leading to toxin production in situ. An important next step is to determine whether the identified m/z + 1 values in our models emerge in samples from other lakes experiencing toxic cyanoHAB events. Because water manager actions are predicated on sensitive and timely detection of cyanotoxins and their bacterial producers, future research that harnesses volatilomes in conjunction with other accessible complex data, including real-time buoy and satellite monitoring, to track and predict cyanoHAB trajectories before, during, and after toxic HAB events, is warranted to limit public exposures and economic hardship.

MATERIALS AND METHODS

Water sample collection

Water samples were collected from three sites on Upper Klamath Lake and one site on its northern arm, Agency Lake, during the months of May–December in 2018 and 2019. Sampling sites were NAL, an agricultural-dominated terrain on the northeastern shore of Agency Lake, two wildlife and recreational areas near the peninsula at Eagle Point in UKL (EPP) and western shore of UKL at Howard Bay (WBR), and a residential area near the dam at the southern end of the lake (PEL) (Fig. 1). Additional samples were collected from canals that drain from UKL for agricultural irrigation (NCA, ADY, MER, and LRI; Fig. 1). Samples were collected by pole from the surface about 2 m from the shore or canal line. Samples for all analyses (VOCs, anions, pigments, particulate C and N, community composition) excepting microcystin concentration were collected in autoclaved 1-L polycarbonate bottles with limited to no headspace. Samples were transported in a cooler to Oregon State University, Corvallis, OR. VOCs were measured and microscopic analysis was conducted within 24 hours of collection. Samples for POC, PON, chlorophyll, and 16S rRNA sequencing were filtered and frozen within 24 hours of collection. Samples for microcystin concentration were collected on site in autoclaved 10-mL glass vials and frozen on arrival at Oregon State University (−20°C) for later analysis. Samples for ion measurement were frozen for later analysis at Oregon State Universities Freshwater IIW Collaboratory.

Environmental parameters

Temperature, pH, and conductivity were measured on site using an Extech pH/temperature meter (Nashua, NH) and YSI 30 Conductivity meter (Yellow Springs, OH), respectively. The anions bromide, fluoride, chloride, nitrite, nitrate, phosphate, and sulfate were measured with a Dionex ICS-1500 Ion Chromatograph Autosampler (Sunnyvale, CA). Data for bromide and fluoride are not shown because only two samples yielded data above the detection limits. Ammonium was measured by UV-Vis spectroscopy after three freeze-thaw cycles (98). POC and PON were determined from three volumes (3–110 mL) filtered onto pre-combusted GF/F filters to create a linear regression, and frozen until analysis by Exeter Analytical EA1 elemental analyzer (99) (Coventry, England). Non-particulate C and N were determined from sample filtrate and subtracted from the filtered samples. The median sample volume of filtrate was re-filtered onto a fourth GF/F filter, frozen, and analyzed with the sample filters. Chlorophyll concentration was measured in triplicate using 2–100 mL of sample filtered onto 25-mm GF/F filters (until green was observed on the filter) and extracted for 24–48 hours at −20°C in 90% acetone. Extract absorption was measured by spectrophotometer (Shimadzu, Kyoto, Japan) and calculated using the equation for cyanobacteria from Ritchie (2006) (100). Microcystin concentrations were measured using Eurofins Abraxis Inc. Microcystins/Nodularins (ADDA) ELISA Kit (Product No. 520011).

Detection of VOCs

Triplicate 100-mL sub-samples were transferred to custom-made 200-mL polycarbonate dynamic stripping chambers with sintered glass frits (2–2.5 µm) at the bases (32). Chambers were placed in an incubator at the sample collection temperature. Samples were stripped of VOCs by flowing synthetic air through a hydrocarbon trap, then a flow controller (Sierra Instruments) set to 50 sccm, and then through the glass frits into the samples. The carrier air with the stripped VOCs was directed into the PTR-TOF-MS (Ionicon, Austria) where the primary ion (H₃O⁺) causes a proton transfer reaction, or soft ionization event, to VOCs having higher proton affinities than 691 kJ/mol, which is the proton affinity of water. VOCs in the mass range of 18–363 a.m.u. were detected at their molecular masses plus one (m/z + 1). Data were collected over 5 minutes. The conditions of the drift tube were 2.1 mbar, 80°C, and 500 V with an E/N value of 125 Td.

VOC data processing

PTR-TOF-MS raw peak data were processed using PTRwid (101). The resulting output yielded tables giving each integrated m/z + 1 peak signal that incorporated a correction for overlapping peaks. PTRwid yields a unified mass list of all m/z + 1 values detected in all lake and canal samples. Known contaminants and internal standards were removed from the list prior to subsequent data processing (Table S4). The first 2.5 minutes of data were removed to account for contaminating air in the tubing and headspace of the stripping chambers. The remaining data were integrated over 2.5–5 minutes. Differences in concentrations of m/z + 1 values and Wald’s test derived P-values were determined using the R package DESeq2 (102), and Benjamini-Hochberg corrected P-values of m/z + 1 values were determined using the EnhancedVolcano package (102). Chemical formulas were assigned using Ionicon PTR Viewer software and the Ionicon integrated database (PTR Viewer software version 3.3), PTR Viewer calculated values (version 3.4.2) or GLOVOCs database for PTR-MS (103) (update November 16, 2020). Some chemical formulas were assigned based on published PTR-MS research on those compounds. The maximum mass shift (difference between the actual mass value and the detected mass value) allowed for the compound assignment was 0.007 a.m.u. as determined by PTR-TOF-MS calculated RMSE.

VOC-based elastic net models predicting microcystin concentration

Two elastic net model types were developed: (i) linear models that predict the continuous outcome of microcystin concentration (M1 and M7) and (ii) logistical models that predict the dichotomous outcome of whether microcystin concentration ≥0.3 ppb (M2 and M8). The logistical models were trained with a binary output that designated a sample as toxic if the microcystin concentration was at or above the 0.3 ppb threshold. M1 and M2 utilized only m/z + 1 values and were trained using the glmnet function in R software (version 4.1.0) on 95% of the samples (total n = 70). M7 and M8 utilized m/z + 1 values and “low-cost” environmental variables (Table 2) and were trained on 95% of all samples for which VOC and “low-cost” environmental data were available (total n = 35). Cross-validation with 10- (M7, M8) or 15 (M1, M2)-folds was used to determine the values of the tuning parameters (Table S2), and hence the strength of regularization in M1, M2, M7, and M8. MSPE and their SD in M1 and M7 were calculated based on the model with the tuning parameter set to the value yielding the minimum mean cross-validated error. AUC and AIC for M2 and M8 were averaged from 10 random iterations of each model. The m/z + 1 values in the final models were selected using the full data sets. The m/z + 1 values retained by the elastic net models were refit using the lm function in R to yield the coefficients in each final model (Table S2). A lower MSPE, lower AIC, and higher AUC are indicative of a preferred model.

“Low cost” and “low + high-cost” regression models predicting microcystin concentration

Four base models were developed to predict microcystin concentration using only environmental parameters (Table 2). Outlier removal and bidirectional stepwise elimination were implemented using the MASS package in R to select the environmental parameters. Microcystin concentration in M3 was modeled by multiple linear regression using the month of collection, the collection site, and “low-cost” environmental parameters. Multiple linear regression model, M5, was based on the month of collection, collection site, and both “low-cost” and “high-cost” environmental parameters (Table 3). Two logistic models were developed based on “low-cost” and “low + high-cost” environmental parameters (M4 and M6, respectively) to predict microcystin concentration ≥0.3 ppb. The predictive performance of these linear regression base models was evaluated using the glmnet (104, 105) package with lambda and alpha values set to zero.

DNA extraction and sequencing

Samples (10–150 mL) were filtered onto 0.2-µm polycarbonate filters and stored at −20°C until DNA extraction using phenol:choloroform:iso-amyl 25:24:1. DNA quality and quantity were determined by NanoDrop 1000¹⁰⁴. The V1–V2 region of the 16S rRNA gene was amplified using 27F (5′-AGAAGAGTTTGATCNTGGCTCAG-3′) and 338 RPL (5′-CWGCCWCCCGTAGGWGT-3′) primers with overhang adaptors according to the Illumina Inc. standard 16S sequencing library preparation protocol. Libraries were created using dual indices and Illumina sequencing adapters with a Nextera XT Index Kit (Illumina Inc.), then pooled in equimolar concentrations and sequenced using Illumina MiSeq (2× 250 PE) in two batches: 54 samples were sequenced at the Center for Quantitative Life Sciences (Oregon State University, Oregon), and 16 samples were sequenced at Molecular Research DNA-RNA Laboratory (Shallowater, TX) (106).

16S rRNA gene amplicon analysis

The 27F (20 bp) and 338RPL (18 bp) primers were removed using CutAdapt, then DADA version 1.2 R package (version 3.6.1), using the SILVA database train version 138, quality-filtered, dereplicated, merged, constructed an ASV table, removed chimeras, and taxonomically assigned the sample reads through the dada2 package (106). Taxonomic assignment through dada2 was used for phylum and class classification. Taxonomic assignment of 16S rRNA sequences within the Cyanobacteria genera was done using Cydrasil and its maximum-likelihood phylogenetic tree constructed of 1,327 Cyanobacteria reference sequences (107). Sequences were aligned using reference alignments constructed using PaPaRa version 2.0, and the alignments were placed using EPA-ng (107). The placements were visualized via the Interactive Tree of Life (iTOL) version 6.5.4, and taxonomy was hand-assigned (Fig. S6) (108). The dominant cyanobacteria were also confirmed by visual identification using light microscopy and morphological characteristics (Fig. S3).

VOC-based models of microbial community composition

Relative abundances of the four most abundant microbial phyla, classes, and cyanobacterial genera (Anabaena, Aphanizomenon, Gloeotrichia, and Microcystis) were determined using the R package phyloseq (109). A total of 12 VOC-based linear regression models were developed to predict microbial relative abundances using elastic net modeling. Models were trained using the glmnet function in R (version 4.1.0) on 95% of the samples (total n = 70). Cross-validation with 15-folds was used to determine the value of the tuning parameter (Table S3) and hence th strength of regularization. MSPE and their SDs were calculated from the final model with the tuning parameter set to the value yielding the minimum mean cross-validated error. The m/z + 1 values in the final models were selected using the full data set. The m/z + 1 values retained by the elastic net models were refit using the lm function in R to yield the coefficients in each final model (Table S3). The R package ComplexHeatMap (110) was used to create Fig. 2 and 6, and Fig. S2, and the R package EnhancedVolcano (111) was used to create Fig. 3.

ACKNOWLEDGMENTS

We are grateful for the funding support from the College of Science SciRIS program and Agricultural Research Foundation at Oregon State University. K.H.H. is especially thankful for funds from an anonymous donor that established the Excellence in Microbiology Faculty Scholar award and supported this research.

We thank Dr. Theo Dreher for assistance with sample collection and Dr. Benjamin Philmus for his insight with mass spectrometry.

L.C., D.J., and K.H.H. designed the study. L.C. collected the data and developed the elastic net models. K.H.H. contributed overall project oversight. D.J. supervised model development. L.C. and K.H.H. wrote the manuscript with input from D.J.

Contributor Information

Kimberly H. Halsey, Email: Kimberly.Halsey@oregonstate.edu.

Rachel Poretsky, University of Illinois at Chicago, Chicago, Illinois, USA .

DATA AVAILABILITY

The authors declare that all data supporting the results of this study are available within the article and corresponding supplemental material. 16S rRNA sequencing data have been deposited to National Center for Biotechnology Information (NCBI accession no. PRJNA922214), and the model code is available on Open Science Framework.

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/msystems.00379-23.

Table S3. msystems.00379-23-s0001.csv.

MSPE, SD, intercept, alpha value, and coefficients for m/z+1 values that were retained in the elastic net models predicting bacterial relative abundance.

Click here for additional data file.^{(6.3KB, csv)}

DOI: 10.1128/msystems.00379-23.SuF1

Supplemental figures and tables. msystems.00379-23-s0002.docx.

Tables S1, S2, and S4; Figures S1-S6.

Click here for additional data file.^{(2.8MB, docx)}

DOI: 10.1128/msystems.00379-23.SuF2

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.

REFERENCES

1. Lee S, Jiang X, Manubolu M, Riedl K, Ludsin SA, Martin JF, Lee J. 2017. Fresh produce and their soils accumulate cyanotoxins from irrigation water: implications for public health and food security. Food Res Int 102:234–245. doi: 10.1016/j.foodres.2017.09.079 [DOI] [PubMed] [Google Scholar]
2. Lee J, Lee S, Jiang X. 2017. Cyanobacterial toxins in freshwater and food: important sources of exposure to humans. Annu Rev Food Sci Technol 8:281–304. doi: 10.1146/annurev-food-030216-030116 [DOI] [PubMed] [Google Scholar]
3. Dreher TW, Collart LP, Mueller RS, Halsey KH, Bildfell RJ, Schreder P, Sobhakumari A, Ferry R. 2019. Anabaena/Dolichospermum as the source of lethal microcystin levels responsible for a large cattle toxicosis event. Toxicon X 1:100003. doi: 10.1016/j.toxcx.2018.100003 [DOI] [PMC free article] [PubMed] [Google Scholar]
4. Massey IY, Yang F, Ding Z, Yang S, Guo J, Tezi C, Al-Osman M, Kamegni RB, Zeng W. 2018. Exposure routes and health effects of microcystins on animals and humans: a mini-review. Toxicon 151:156–162. doi: 10.1016/j.toxicon.2018.07.010 [DOI] [PubMed] [Google Scholar]
5. Badar M, Batool F, Khan SS, Khokhar I, Qamar MK, Yasir C. 2017. Effects of microcystins toxins contaminated drinking water on hepatic problems in animals (cows and buffalos) and toxins removal chemical method. Buffalo Bulletin 36:43–56. [Google Scholar]
6. Weralupitiya C, Wanigatunge RP, Gunawardana D, Vithanage M, Magana-Arachchi D. 2022. Cyanotoxins uptake and accumulation in crops: phytotoxicity and implications on human health. Toxicon 211:21–35. doi: 10.1016/j.toxicon.2022.03.003 [DOI] [PubMed] [Google Scholar]
7. Bouaïcha N, Corbel S.. 2016. Cyanobacterial toxins emerging contaminants in soils: a review of sources, fate and Impacts on ecosystems, plants and animal and human health, Ch 6. In Larramendy ML, Soloneski S (ed), Soil Contamination - Current Consequences and Further Solutions, InTechOpen, Rijeka. [Google Scholar]
8. Holland A, Kinnear S. 2013. Interpreting the possible ecological role(s) of cyanotoxins: compounds for competitive advantage and/or physiological aide. Mar Drugs 11:2239–2258. doi: 10.3390/md11072239 [DOI] [PMC free article] [PubMed] [Google Scholar]
9. Christoffersen K, Lyck S, Winding A. 2002. Microbial activity and bacterial community structure during degradation of microcystins. Aquat Microb Ecol 27:125–136. doi: 10.3354/ame027125 [DOI] [Google Scholar]
10. Anderson DM, Hoagland P, Kaoru Y, White AW. 2000. Estimated annual economic impacts from harmful algal blooms (HABs) in the United States. Woods Hole Oceanographic Institution. doi: 10.1575/1912/96 [DOI] [Google Scholar]
11. Watson SB, Whitton BA, Higgins SN, Paerl HW, Brooks BW, Wehr JD. 2015. Chapter Ch 20, Harmful algal blooms, p 873–920. In Freshwater algae of North America. Academic Press, Boston. [Google Scholar]
12. Huisman J, Codd GA, Paerl HW, Ibelings BW, Verspagen JMH, Visser PM. 2018. Cyanobacterial blooms. Nat Rev Microbiol 16:471–483. doi: 10.1038/s41579-018-0040-1 [DOI] [PubMed] [Google Scholar]
13. Paerl HW, Huisman J. 2008. Climate: blooms like it hot. Science 320:57–58. doi: 10.1126/science.1155398 [DOI] [PubMed] [Google Scholar]
14. O’Neil JM, Davis TW, Burford MA, Gobler CJ. 2012. The rise of harmful cyanobacteria blooms: the potential roles of eutrophication and climate change. Harmful Algae 14:313–334. doi: 10.1016/j.hal.2011.10.027 [DOI] [Google Scholar]
15. Jöhnk KD, Huisman J, Sharples J, Sommeijer B, Visser PM, Stroom JM. 2008. Summer heatwaves promote blooms of harmful cyanobacteria. Glob Chang Biol 14:495–512. doi: 10.1111/j.1365-2486.2007.01510.x [DOI] [Google Scholar]
16. Anneville O, Domaizon I, Kerimoglu O, Rimet F, Jacquet S. 2015. “Blue-green algae in a “greenhouse century”? new insights from field data on climate change impacts on cyanobacteria abundance”. Ecosystems 18:441–458. doi: 10.1007/s10021-014-9837-6 [DOI] [Google Scholar]
17. Schmidt JR, Wilhelm SW, Boyer GL. 2014. The fate of microcystins in the environment and challenges for monitoring. Toxins (Basel) 6:3354–3387. doi: 10.3390/toxins6123354 [DOI] [PMC free article] [PubMed] [Google Scholar]
18. Pacheco ABF, Guedes IA, Azevedo SMFO. 2016. Is qPCR a reliable indicator of cyanotoxin risk in freshwater. Toxins (Basel) 8:172. doi: 10.3390/toxins8060172 [DOI] [PMC free article] [PubMed] [Google Scholar]
19. Massey IY, Wu P, Wei J, Luo J, Ding P, Wei H, Yang F. 2020. A mini-review on detection methods of microcystins. Toxins (Basel) 12:641. doi: 10.3390/toxins12100641 [DOI] [PMC free article] [PubMed] [Google Scholar]
20. Jones MR, Pinto E, Torres MA, Dörr F, Mazur-Marzec H, Szubert K, Tartaglione L, Dell’Aversano C, Miles CO, Beach DG, McCarron P, Sivonen K, Fewer DP, Jokela J, Janssen EM-L. 2021. CyanoMetDB, a comprehensive public database of secondary metabolites from cyanobacteria. Water Res 196:117017. doi: 10.1016/j.watres.2021.117017 [DOI] [PubMed] [Google Scholar]
21. Almuhtaram H, Kibuye FA, Ajjampur S, Glover CM, Hofmann R, Gaget V, Owen C, Wert EC, Zamyadi A. 2021. State of knowledge on early warning tools for cyanobacteria detection. Ecological Indicators 133:108442. doi: 10.1016/j.ecolind.2021.108442 [DOI] [Google Scholar]
22. Alivisatos AP, Blaser MJ, Brodie EL, Chun M, Dangl JL, Donohue TJ, Dorrestein PC, Gilbert JA, Green JL, Jansson JK, Knight R, Maxon ME, McFall-Ngai MJ, Miller JF, Pollard KS, Ruby EG, Taha SA, Unified Microbiome Initiative Consortium . 2015. A unified initiative to harness earth’s microbiomes. Science 350:507–508. doi: 10.1126/science.aac8480 [DOI] [PubMed] [Google Scholar]
23. Goutman SA, Boss J, Guo K, Alakwaa FM, Patterson A, Kim S, Savelieff MG, Hur J, Feldman EL. 2020. Untargeted metabolomics yields insight into ALS disease mechanisms. J Neurol Neurosurg Psychiatry 91:1329–1338. doi: 10.1136/jnnp-2020-323611 [DOI] [PMC free article] [PubMed] [Google Scholar]
24. Zhang H, Wang J, Sun Z, Zurada JM, Pal NR. 2020. Feature selection for neural networks using group lasso regularization. IEEE Trans Knowl Data Eng 32:659–673. doi: 10.1109/TKDE.2019.2893266 [DOI] [Google Scholar]
25. Pozzer AC, Gómez PA, Weiss J. 2022. Volatile organic compounds in aquatic ecosystems – detection, origin, significance and applications. Sci Total Environ 838:156155. doi: 10.1016/j.scitotenv.2022.156155 [DOI] [PubMed] [Google Scholar]
26. Weisskopf L, Schulz S, Garbeva P. 2021. Microbial volatile organic compounds in intra-kingdom and inter-kingdom interactions. Nat Rev Microbiol 19:391–404. doi: 10.1038/s41579-020-00508-1 [DOI] [PubMed] [Google Scholar]
27. Amavizca E, Bashan Y, Ryu C-M, Farag MA, Bebout BM, de-Bashan LE. 2017. Enhanced performance of the Microalga chlorella sorokiniana remotely induced by the plant growth-promoting bacteria Azospirillum brasilense and Bacillus pumilus. Sci Rep 7:41310. doi: 10.1038/srep41310 [DOI] [PMC free article] [PubMed] [Google Scholar]
28. Fink P. 2007. Ecological functions of volatile organic compounds in aquatic systems. Mar Freshw Behav Physiol 40:155–168. doi: 10.1080/10236240701602218 [DOI] [Google Scholar]
29. Xu Q, Yang L, Yang W, Bai Y, Hou P, Zhao J, Zhou L, Zuo Z. 2017. Volatile organic compounds released from Microcystis flos-aquae under nitrogen sources and their toxic effects on Chlorella Vulgaris. Ecotoxicology and Environmental Safety 135:191–200. doi: 10.1016/j.ecoenv.2016.09.027 [DOI] [PubMed] [Google Scholar]
30. Zuo Z-J, Zhu Y-R, Bai Y-L, Wang Y. 2012. Volatile communication between Chlamydomonas reinhardtii cells under salt stress. Biochemical Systematics and Ecology 40:19–24. doi: 10.1016/j.bse.2011.09.007 [DOI] [Google Scholar]
31. Schulz-Bohm K, Geisen S, Wubs ERJ, Song C, de Boer W, Garbeva P. 2017. The prey’s scent – volatile organic compound mediated interactions between soil bacteria and their protist predators. ISME J 11:817–820. doi: 10.1038/ismej.2016.144 [DOI] [PMC free article] [PubMed] [Google Scholar]
32. Halsey KH, Giovannoni SJ, Graus M, Zhao Y, Landry Z, Thrash JC, Vergin KL, de Gouw J. 2017. Biological cycling of volatile organic carbon by phytoplankton and bacterioplankton. Limnol Oceanogr 62:2650–2661. doi: 10.1002/lno.10596 [DOI] [Google Scholar]
33. Moore ER, Weaver AJ, Davis EW, Giovannoni SJ, Halsey KH. 2022. Metabolism of key atmospheric volatile organic compounds by the marine heterotrophic bacterium Pelagibacter HTCC1062 (SAR11). Environ Microbiol 24:212–222. doi: 10.1111/1462-2920.15837 [DOI] [PMC free article] [PubMed] [Google Scholar]
34. Aneja VP, Schlesinger WH, Erisman JW. 2009. Effects of agriculture upon the air quality and climate: research, policy, and regulations. Environ Sci Technol 43:4234–4240. doi: 10.1021/es8024403 [DOI] [PubMed] [Google Scholar]
35. Ozaki K, Ohta A, Iwata C, Horikawa A, Tsuji K, Ito E, Ikai Y, Harada K-I. 2008. Lysis of cyanobacteria with volatile organic compounds. Chemosphere 71:1531–1538. doi: 10.1016/j.chemosphere.2007.11.052 [DOI] [PubMed] [Google Scholar]
36. Ye C, Yang Y, Xu Q, Ying B, Zhang M, Gao B, Ni B, Yakefu Z, Bai Y, Zuo Z. 2018. Volatile organic compound emissions from Microcystis aeruginosa under different phosphorus sources and concentrations. Phycological Res 66:15–22. doi: 10.1111/pre.12201 [DOI] [Google Scholar]
37. Zuo Z, Yang Y, Xu Q, Yang W, Zhao J, Zhou L. 2018. Effects of phosphorus sources on volatile organic compound emissions from Microcystis flos-aquae and their toxic effects on Chlamydomonas reinhardtii. Environ Geochem Health 40:1283–1298. doi: 10.1007/s10653-017-0055-y [DOI] [PubMed] [Google Scholar]
38. Ikawa M, Sasner JJ, Haney JF. 2001. Activity of cyanobacterial and algal odor compounds found in Lake waters on green alga Chlorella pyrenoidosa growth. Hydrobiologia 443:19–22. doi: 10.1023/A:1017535801766 [DOI] [Google Scholar]
39. Yamashita R, Bober B, Kanei K, Arii S, Tsuji K, Harada K-I. 2020. Analytical technique optimization on the detection of β-cyclocitral in Microcystis species. Molecules 25:832. doi: 10.3390/molecules25040832 [DOI] [PMC free article] [PubMed] [Google Scholar]
40. Zuo Z, Zhu Y, Bai Y, Wang Y. 2012. Acetic acid-induced programmed cell death and release of volatile organic compounds in Chlamydomonas reinhardtii. Plant Physiol Biochem 51:175–184. doi: 10.1016/j.plaphy.2011.11.003 [DOI] [PubMed] [Google Scholar]
41. Liu M, Wu T, Zhao X, Zan F, Yang G, Miao Y. 2021. Cyanobacteria blooms potentially enhance volatile organic compound (VOC) emissions from a Eutrophic Lake: field and experimental evidence. Environ Res 202:111664. doi: 10.1016/j.envres.2021.111664 [DOI] [PubMed] [Google Scholar]
42. Zuo Z, Yang L, Chen S, Ye C, Han Y, Wang S, Ma Y. 2018. Effects of nitrogen nutrients on the volatile organic compound emissions from Microcystis aeruginosa. Ecotoxicol Environ Saf 161:214–220. doi: 10.1016/j.ecoenv.2018.05.095 [DOI] [PubMed] [Google Scholar]
43. Bižić M, Klintzsch T, Ionescu D, Hindiyeh MY, Günthel M, Muro-Pastor AM, Eckert W, Urich T, Keppler F, Grossart H-P. 2020. Aquatic and terrestrial cyanobacteria produce methane. Sci Adv 6:eaax5343. doi: 10.1126/sciadv.aax5343 [DOI] [PMC free article] [PubMed] [Google Scholar]
44. Santos AB, Fernandes AS, Wagner R, Jacob-Lopes E, Zepka LQ. 2016. Biogeneration of volatile organic compounds produced by Phormidium autumnale in heterotrophic bioreactor. J Appl Phycol 28:1561–1570. doi: 10.1007/s10811-015-0740-0 [DOI] [Google Scholar]
45. Lindberg P, Park S, Melis A. 2010. Engineering a platform for photosynthetic isoprene production in cyanobacteria, using Synechocystis as the model organism. Metab Eng 12:70–79. doi: 10.1016/j.ymben.2009.10.001 [DOI] [PubMed] [Google Scholar]
46. O’Dowd CD, de Leeuw G. 2007. Marine aerosol production: a review of the current knowledge. Philos Trans A Math Phys Eng Sci 365:1753–1774. doi: 10.1098/rsta.2007.2043 [DOI] [PubMed] [Google Scholar]
47. Moore ER, Davie-Martin CL, Giovannoni SJ, Halsey KH. 2020. Pelagibacter metabolism of diatom‐derived volatile organic compounds imposes an energetic tax on photosynthetic carbon fixation. Environ Microbiol 22:1720–1733. doi: 10.1111/1462-2920.14861 [DOI] [PubMed] [Google Scholar]
48. Kieber DJ, McDaniel J, Mopper K. 1989. Photochemical source of biological substrates in sea water: implications for carbon cycling. Nature 341:637–639. doi: 10.1038/341637a0 [DOI] [Google Scholar]
49. Vuorio E, Thiel K, Fitzpatrick D, Huokko T, Kämäräinen J, Dandapani H, Aro E-M, Kallio P. 2021. Hydrocarbon desaturation in cyanobacterial thylakoid membranes is linked with acclimation to suboptimal growth temperatures. Front Microbiol 12:781864. doi: 10.3389/fmicb.2021.781864 [DOI] [PMC free article] [PubMed] [Google Scholar]
50. Lea-Smith DJ, Ortiz-Suarez ML, Lenn T, Nürnberg DJ, Baers LL, Davey MP, Parolini L, Huber RG, Cotton CAR, Mastroianni G, Bombelli P, Ungerer P, Stevens TJ, Smith AG, Bond PJ, Mullineaux CW, Howe CJ. 2016. Hydrocarbons are essential for optimal cell size, division, and growth of cyanobacteria. Plant Physiol 172:1928–1940. doi: 10.1104/pp.16.01205 [DOI] [PMC free article] [PubMed] [Google Scholar]
51. Qiao Y, Wang W, Lu X. 2020. High light induced alka(e)ne biodegradation for lipid and redox homeostasis in cyanobacteria. Front. Microbiol 11:1659. doi: 10.3389/fmicb.2020.01659 [DOI] [PMC free article] [PubMed] [Google Scholar]
52. Berla BM, Saha R, Maranas CD, Pakrasi HB. 2015. Cyanobacterial alkanes modulate photosynthetic cyclic electron flow to assist growth under cold stress. Sci Rep 5:14894. doi: 10.1038/srep14894 [DOI] [PMC free article] [PubMed] [Google Scholar]
53. Oregon Health Authority. Accession May 15, 2023. https://www.oregon.gov/oha/ph/healthyenvironments/recreation/harmfulalgaeblooms/pages/blue-greenalgaeadvisories.aspx. [Google Scholar]
54. United States Environmental Protection Agency. Accession May 15, 2023. https://www.epa.gov/cyanohabs/epa-drinking-water-health-advisories-cyanotoxins. [Google Scholar]
55.Agency U.S. Environmental ProtectionHealth and ecological criteria division. Washington, DC. 2019. Recommended human health recreational ambient water quality criteria or swimming advisories for microcystins and cylindrospermopsin. EPA Document Number: 822-R-19-001.
56. Carmichael WW, Drapeau C, Anderson DM. 2000. Harvesting of Aphanizomenon flos-aquae ralfs ex born. J Appl Phycol 12:585–595. doi: 10.1023/A:1026506713560 [DOI] [Google Scholar]
57. Gugger M, Lyra C, Henriksen P, Couté A, Humbert J-F, Sivonen K. 2002. Phylogenetic comparison of the cyanobacterial genera Anabaena and Aphanizomenon. Int J Syst Evol Microbiol 52:1867–1880. doi: 10.1099/00207713-52-5-1867 [DOI] [PubMed] [Google Scholar]
58. Dreher TW, Davis EW, Mueller RS. 2021. Complete genomes derived by directly sequencing freshwater bloom populations emphasize the significance of the genus level ADA clade within the Nostocales. Harmful Algae 103:102005. doi: 10.1016/j.hal.2021.102005 [DOI] [PubMed] [Google Scholar]
59. Rousso BZ, Bertone E, Stewart R, Hamilton DP. 2020. A systematic literature review of forecasting and predictive models for cyanobacteria blooms in freshwater lakes. Water Res 182:115959. doi: 10.1016/j.watres.2020.115959 [DOI] [PubMed] [Google Scholar]
60. Madsen R, Lundstedt T, Trygg J. 2010. Chemometrics in metabolomics—a review in human disease diagnosis. Anal Chim Acta 659:23–33. doi: 10.1016/j.aca.2009.11.042 [DOI] [PubMed] [Google Scholar]
61. Mallick H, Franzosa EA, Mclver LJ, Banerjee S, Sirota-Madi A, Kostic AD, Clish CB, Vlamakis H, Xavier RJ, Huttenhower C. 2019. Predictive metabolomic profiling of microbial communities using amplicon or metagenomic sequences. Nat Commun 10:3136. doi: 10.1038/s41467-019-10927-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
62. Morrow AL, Lagomarcino AJ, Schibler KR, Taft DH, Yu Z, Wang B, Altaye M, Wagner M, Gevers D, Ward DV, Kennedy MA, Huttenhower C, Newburg DS. 2013. Early microbial and metabolomic signatures predict later onset of necrotizing enterocolitis in preterm infants. Microbiome 1:13. doi: 10.1186/2049-2618-1-13 [DOI] [PMC free article] [PubMed] [Google Scholar]
63. Wilson RM, Tfaily MM, Kolton M, Johnston ER, Petro C, Zalman CA, Hanson PJ, Heyman HM, Kyle JE, Hoyt DW, Eder EK, Purvine SO, Kolka RK, Sebestyen SD, Griffiths NA, Schadt CW, Keller JK, Bridgham SD, Chanton JP, Kostka JE. 2021. Soil metabolome response to whole-ecosystem warming at the spruce and peatland responses under changing environments experiment. Proc Natl Acad Sci U S A 118:e2004192118. doi: 10.1073/pnas.2004192118 [DOI] [PMC free article] [PubMed] [Google Scholar]
64. Lawson CA, Possell M, Seymour JR, Raina J-B, Suggett DJ. 2019. Coral endosymbionts (Symbiodiniaceae) emit species-specific volatilomes that shift when exposed to thermal stress. Sci Rep 9:17395. doi: 10.1038/s41598-019-53552-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
65. Meskhidze N, Sabolis A, Reed R, Kamykowski D. 2015. Quantifying environmental stress-induced emissions of algal isoprene and monoterpenes using laboratory measurements. Biogeosciences 12:637–651. doi: 10.5194/bg-12-637-2015 [DOI] [Google Scholar]
66. Dani KGS, Loreto F. 2017. Trade-off between dimethyl sulfide and isoprene emissions from marine phytoplankton. Trends Plant Sci 22:361–372. doi: 10.1016/j.tplants.2017.01.006 [DOI] [PubMed] [Google Scholar]
67. Noecker C, Eng A, Srinivasan S, Theriot CM, Young VB, Jansson JK, Fredricks DN, Borenstein E, Sanchez LM. 2016. Metabolic model-based integration of microbiome taxonomic and metabolomic profiles elucidates mechanistic links between ecological and metabolic variation. mSystems 1:e00013-15. doi: 10.1128/mSystems.00013-15 [DOI] [PMC free article] [PubMed] [Google Scholar]
68. Morton JT, Aksenov AA, Nothias LF, Foulds JR, Quinn RA, Badri MH, Swenson TL, Van Goethem MW, Northen TR, Vazquez-Baeza Y, Wang M, Bokulich NA, Watters A, Song SJ, Bonneau R, Dorrestein PC, Knight R. 2019. Learning representations of microbe–metabolite interactions. Nat Methods 16:1306–1314. doi: 10.1038/s41592-019-0616-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
69. Quinn TP, Erb I. 2021. Examining microbe–metabolite correlations by linear methods. Nat Methods 18:37–39. doi: 10.1038/s41592-020-01006-1 [DOI] [PubMed] [Google Scholar]
70. Morton JT, McDonald D, Aksenov AA, Nothias LF, Foulds JR, Quinn RA, Badri MH, Swenson TL, Van Goethem MW, Northen TR, Vazquez-Baeza Y, Wang M, Bokulich NA, Watters A, Song SJ, Bonneau R, Dorrestein PC, Knight R. 2021. Reply to: examining microbe–metabolite correlations by linear methods. Nat Methods 18:40–41. doi: 10.1038/s41592-020-01007-0 [DOI] [PubMed] [Google Scholar]
71. Agger SA, Lopez-Gallego F, Hoye TR, Schmidt-Dannert C. 2008. Identification of sesquiterpene synthases from Nostoc punctiforme PCC 73102 and Nostoc sp. strain PCC 7120. J Bacteriol 190:6084–6096. doi: 10.1128/JB.00759-08 [DOI] [PMC free article] [PubMed] [Google Scholar]
72. Merel S, Walker D, Chicana R, Snyder S, Baurès E, Thomas O. 2013. State of knowledge and concerns on cyanobacterial blooms and cyanotoxins. Environ Int 59:303–327. doi: 10.1016/j.envint.2013.06.013 [DOI] [PubMed] [Google Scholar]
73. Muñoz-González C, Canon F, Feron G, Guichard E, Pozo-Bayón M. 2019. Assessment wine aroma persistence by using an in vivo PTR-ToF-MS approach and its relationship with salivary parameters. Molecules 24:1277. doi: 10.3390/molecules24071277 [DOI] [PMC free article] [PubMed] [Google Scholar]
74. Zhang K, Lin TF, Zhang T, Li C, Gao N. 2013. Characterization of typical taste and odor compounds formed by Microcystis aeruginosa. J Environ Sci (China) 25:1539–1548. doi: 10.1016/s1001-0742(12)60232-0 [DOI] [PubMed] [Google Scholar]
75. Shao J, Peng L, Luo S, Yu G, Gu J, Lin S, Li R. 2013. First report on the allelopathic effect of Tychonema bourrellyi (cyanobacteria) against Microcystis aeruginosa (cyanobacteria). J Appl Phycol 25:1567–1573. doi: 10.1007/s10811-012-9969-z [DOI] [Google Scholar]
76. Liu J, Yu Q, Ye B, Zhu K, Yin J, Zheng T, Xu S, Sun Q, Li Y, Zuo Z. 2021. Programmed cell death of Chlamydomonas reinhardtii induced by three cyanobacterial volatiles Β-ionone, limonene and longifolene. Sci Total Environ 762:144539. doi: 10.1016/j.scitotenv.2020.144539 [DOI] [PubMed] [Google Scholar]
77. Shao J, Xu Y, Wang Z, Jiang Y, Yu G, Peng X, Li R. 2011. Elucidating the toxicity targets of Β-ionone on photosynthetic system of Microcystis aeruginosa NIES-843 (cyanobacteria). Aquat Toxicol 104:48–55. doi: 10.1016/j.aquatox.2011.03.014 [DOI] [PubMed] [Google Scholar]
78. Walsh K, Jones GJ, Dunstan RH. 1998. Effect of high irradiance and iron on volatile odour compounds in the cyanobacterium Microcystis aeruginosa. Phytochemistry 49:1227–1239. doi: 10.1016/s0031-9422(97)00943-6 [DOI] [PubMed] [Google Scholar]
79. Suurnäkki S, Gomez-Saez GV, Rantala-Ylinen A, Jokela J, Fewer DP, Sivonen K. 2015. Identification of geosmin and 2-methylisoborneol in cyanobacteria and molecular detection methods for the producers of these compounds. Water Res 68:56–66. doi: 10.1016/j.watres.2014.09.037 [DOI] [PubMed] [Google Scholar]
80. Hellweger FL, Schampera C, Martin RM, Eigemann F, Smith DJ, Dick GJ, Wilhelm SW. 2022. Models predict planned phosphorus load reduction will make lake erie more toxic. Science 378:eade2277. doi: 10.1126/science.ade2277 [DOI] [PubMed] [Google Scholar]
81. Höckelmann C, Jüttner F. 2004. Volatile organic compound (VOC) analysis and sources of limonene, cyclohexanone and straight chain aldehydes in axenic cultures of Calothrix and Plectonema. Water Sci Technol 49:47–54. [PubMed] [Google Scholar]
82. Yang R, Zhu L, Li T, Zhu L, Ye Z, Zhang D. 2021. Photosynthetic conversion of CO2 into pinene using engineered Synechococcus sp. Front Bioeng Biotechnol 9. doi: 10.3389/fbioe.2021.779437 [DOI] [PMC free article] [PubMed] [Google Scholar]
83. Pattanaik B, Lindberg P. 2015. Terpenoids and their biosynthesis in cyanobacteria. Life (Basel) 5:269–293. doi: 10.3390/life5010269 [DOI] [PMC free article] [PubMed] [Google Scholar]
84. Zhao J, Yang L, Zhou L, Bai Y, Wang B, Hou P, Xu Q, Yang W, Zuo Z. 2016. Inhibitory effects of Eucalyptol and limonene on the photosynthetic abilities in Chlorella vulgaris (Chlorophyceae). Phycologia 55:696–702. doi: 10.2216/16-38.1 [DOI] [Google Scholar]
85. Hu X, Liu Y, Zeng G, Hu X, Wang Y, Zeng X. 2014. Effects of limonene stress on the growth of and microcystin release by the freshwater cyanobacterium Microcystis aeruginosa FACHB-905. Ecotoxicol Environ Saf 105:121–127. doi: 10.1016/j.ecoenv.2014.01.023 [DOI] [PubMed] [Google Scholar]
86. Ghanbari J, Khajoei-Nejad G, Erasmus SW, van Ruth SM. 2019. Identification and characterisation of volatile fingerprints of saffron stigmas and petals using PTR-TOF-MS: influence of nutritional treatments and corm provenance. Ind Crops and Prod 141:111803. doi: 10.1016/j.indcrop.2019.111803 [DOI] [Google Scholar]
87. Jalliffier-Merlon E, Marty J-C, Denant V, Saliot A. 1991. Phytoplanktonic sources of volatile aldehydes in the river Rhône estuary. Estuarine Coastal and Shelf Science 32:463–482. doi: 10.1016/0272-7714(91)90035-A [DOI] [Google Scholar]
88. Coates RC, Podell S, Korobeynikov A, Lapidus A, Pevzner P, Sherman DH, Allen EE, Gerwick L, Gerwick WH. 2014. Characterization of cyanobacterial hydrocarbon composition and distribution of biosynthetic pathways. PLoS One 9:e85140. doi: 10.1371/journal.pone.0085140 [DOI] [PMC free article] [PubMed] [Google Scholar]
89. Parveen H, Yazdani SS. 2022. Insights into cyanobacterial alkane biosynthesis. J Ind Microbiol Biotechnol 49:kuab075. doi: 10.1093/jimb/kuab075 [DOI] [PMC free article] [PubMed] [Google Scholar]
90. Xue G, Fiedler AT, Martinho M, Münck E, Que L. 2008. Insights into the P-to-Q conversion in the catalytic cycle of methane monooxygenase from a synthetic model system. Proc Natl Acad Sci USA 105:20615–20620. doi: 10.1073/pnas.0808512105 [DOI] [Google Scholar]
91. Mitschke J, Vioque A, Haas F, Hess WR, Muro-Pastor AM. 2011. Dynamics of transcriptional start site selection during nitrogen stress-induced cell differentiation in Anabaena sp. PCC7120. Proc Natl Acad Sci U S A 108:20130–20135. doi: 10.1073/pnas.1112724108 [DOI] [PMC free article] [PubMed] [Google Scholar]
92. Hernández-Prieto MA, Semeniuk TA, Giner-Lamia J, Futschik ME. 2016. The transcriptional landscape of the photosynthetic model cyanobacterium Synechocystis sp. PCC6803. Sci Rep 6:22168. doi: 10.1038/srep22168 [DOI] [PMC free article] [PubMed] [Google Scholar]
93. Dick GJ, Duhaime MB, Evans JT, Errera RM, Godwin CM, Kharbush JJ, Nitschky HS, Powers MA, Vanderploeg HA, Schmidt KC, Smith DJ, Yancey CE, Zwiers CC, Denef VJ. 2021. The genetic and ecophysiological diversity of Microcystis. Environ Microbiol 23:7278–7313. doi: 10.1111/1462-2920.15615 [DOI] [PubMed] [Google Scholar]
94. Zilliges Y, Kehr J-C, Meissner S, Ishida K, Mikkat S, Hagemann M, Kaplan A, Börner T, Dittmann E. 2011. The cyanobacterial hepatotoxin microcystin binds to proteins and increases the fitness of Microcystis under oxidative stress conditions. PLoS One 6:e17615. doi: 10.1371/journal.pone.0017615 [DOI] [PMC free article] [PubMed] [Google Scholar]
95. Dziallas C, Grossart H-P. 2011. Increasing oxygen radicals and water temperature select for toxic Microcystis sp. PLoS One 6:e25569. doi: 10.1371/journal.pone.0025569 [DOI] [PMC free article] [PubMed] [Google Scholar]
96. Díez-Quijada L, Prieto AI, Guzmán-Guillén R, Jos A, Cameán AM. 2019. Occurrence and toxicity of microcystin congeners other than MC-LR and MC-RR: a review. Food Chem Toxicol 125:106–132. doi: 10.1016/j.fct.2018.12.042 [DOI] [PubMed] [Google Scholar]
97. Van de Waal DB, Verspagen JMH, Lürling M, Van Donk E, Visser PM, Huisman J. 2009. The ecological stoichiometry of toxins produced by harmful cyanobacteria: an experimental test of the carbon-nutrient balance hypothesis. Ecol Lett 12:1326–1335. doi: 10.1111/j.1461-0248.2009.01383.x [DOI] [PubMed] [Google Scholar]
98. Sparks DL, Page AL, Helmke PA, Loeppert RH. (ed) 2020. Methods of Soil Anaylsis, Part 3: Chemical Methods.Wiley.
99. Halsey KH, Milligan AJ, Behrenfeld MJ. 2010. Physiological optimization underlies growth rate-independent chlorophyll-specific gross and net primary production. Photosynth Res 103:125–137. doi: 10.1007/s11120-009-9526-z [DOI] [PubMed] [Google Scholar]
100. Ritchie RJ. 2006. Consistent sets of spectrophotometric chlorophyll equations for acetone, methanol and ethanol solvents. Photosynth Res 89:27–41. doi: 10.1007/s11120-006-9065-9 [DOI] [PubMed] [Google Scholar]
101. Holzinger R. 2015. PTRwid: a new widget tool for processing PTR-TOF-MS data. Atmos Meas Tech 8:3903–3922. doi: 10.5194/amt-8-3903-2015 [DOI] [Google Scholar]
102. Love MI, Huber W, Anders S. 2014. Moderated estimation of fold change and dispersion for RNA-Seq data with DESeq2. Genome Biol 15:550. doi: 10.1186/s13059-014-0550-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
103. Yáñez-Serrano AM, Filella I, LLusià J, Gargallo-Garriga A, Granda V, Bourtsoukidis E, Williams J, Seco R, Cappellin L, Werner C, de Gouw J, Peñuelas J. 2021. GLOVOCS - master compound assignment guide for proton transfer reaction mass spectrometry users. Atmospheric Environment 244:117929. doi: 10.1016/j.atmosenv.2020.117929 [DOI] [Google Scholar]
104. Friedman JH, Hastie T, Tibshirani R. 2010. Regularization paths for generalized linear regression models via coordinate descent. J. Stat. Soft 33:1–22. doi: 10.18637/jss.v033.i01 [DOI] [PMC free article] [PubMed] [Google Scholar]
105. Kenneth Tay J, Narashimhan B, Hastie T. 2023. Elastic net regularization paths for all generalized linear models. J Stat Soft 108:1–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
106. Bolaños LM, Karp-Boss L, Choi CJ, Worden AZ, Graff JR, Haëntjens N, Chase AP, Della Penna A, Gaube P, Morison F, Menden-Deuer S, Westberry TK, O’Malley RT, Boss E, Behrenfeld MJ, Giovannoni SJ. 2020. Small phytoplankton dominate western North Atlantic biomass. ISME J 14:1663–1674. doi: 10.1038/s41396-020-0636-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
107. Roush D, Giraldo-Silva A, Garcia-Pichel F. 2021. Cydrasil 3, a curated 16S rRNA gene reference package and web app for cyanobacterial phylogenetic placement. Sci Data 8:230. doi: 10.1038/s41597-021-01015-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
108. Letunic I, Bork P. 2021. Interactive tree of life (iTOL) v5: an online tool for phylogenetic tree display and annotation. Nucleic Acids Res 49:W293–W296. doi: 10.1093/nar/gkab301 [DOI] [PMC free article] [PubMed] [Google Scholar]
109. McMurdie PJ, Holmes S. 2013. phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS One 8:e61217. doi: 10.1371/journal.pone.0061217 [DOI] [PMC free article] [PubMed] [Google Scholar]
110. Gu Z, Eils R, Schlesner M. 2016. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics 32:2847–2849. doi: 10.1093/bioinformatics/btw313 [DOI] [PubMed] [Google Scholar]
111. Blighe K, Rana S, Lewis M. 2021. Enhancedvolcano: publication-ready volcano plots with enhanced colouring and labeling. Available from: https://github.com/kevinblighe/EnhancedVolcano

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Table S3. msystems.00379-23-s0001.csv.

MSPE, SD, intercept, alpha value, and coefficients for m/z+1 values that were retained in the elastic net models predicting bacterial relative abundance.

Click here for additional data file.^{(6.3KB, csv)}

DOI: 10.1128/msystems.00379-23.SuF1

Supplemental figures and tables. msystems.00379-23-s0002.docx.

Tables S1, S2, and S4; Figures S1-S6.

Click here for additional data file.^{(2.8MB, docx)}

DOI: 10.1128/msystems.00379-23.SuF2

Data Availability Statement

[B1] 1. Lee S, Jiang X, Manubolu M, Riedl K, Ludsin SA, Martin JF, Lee J. 2017. Fresh produce and their soils accumulate cyanotoxins from irrigation water: implications for public health and food security. Food Res Int 102:234–245. doi: 10.1016/j.foodres.2017.09.079 [DOI] [PubMed] [Google Scholar]

[B2] 2. Lee J, Lee S, Jiang X. 2017. Cyanobacterial toxins in freshwater and food: important sources of exposure to humans. Annu Rev Food Sci Technol 8:281–304. doi: 10.1146/annurev-food-030216-030116 [DOI] [PubMed] [Google Scholar]

[B3] 3. Dreher TW, Collart LP, Mueller RS, Halsey KH, Bildfell RJ, Schreder P, Sobhakumari A, Ferry R. 2019. Anabaena/Dolichospermum as the source of lethal microcystin levels responsible for a large cattle toxicosis event. Toxicon X 1:100003. doi: 10.1016/j.toxcx.2018.100003 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B4] 4. Massey IY, Yang F, Ding Z, Yang S, Guo J, Tezi C, Al-Osman M, Kamegni RB, Zeng W. 2018. Exposure routes and health effects of microcystins on animals and humans: a mini-review. Toxicon 151:156–162. doi: 10.1016/j.toxicon.2018.07.010 [DOI] [PubMed] [Google Scholar]

[B5] 5. Badar M, Batool F, Khan SS, Khokhar I, Qamar MK, Yasir C. 2017. Effects of microcystins toxins contaminated drinking water on hepatic problems in animals (cows and buffalos) and toxins removal chemical method. Buffalo Bulletin 36:43–56. [Google Scholar]

[B6] 6. Weralupitiya C, Wanigatunge RP, Gunawardana D, Vithanage M, Magana-Arachchi D. 2022. Cyanotoxins uptake and accumulation in crops: phytotoxicity and implications on human health. Toxicon 211:21–35. doi: 10.1016/j.toxicon.2022.03.003 [DOI] [PubMed] [Google Scholar]

[B7] 7. Bouaïcha N, Corbel S.. 2016. Cyanobacterial toxins emerging contaminants in soils: a review of sources, fate and Impacts on ecosystems, plants and animal and human health, Ch 6. In Larramendy ML, Soloneski S (ed), Soil Contamination - Current Consequences and Further Solutions, InTechOpen, Rijeka. [Google Scholar]

[B8] 8. Holland A, Kinnear S. 2013. Interpreting the possible ecological role(s) of cyanotoxins: compounds for competitive advantage and/or physiological aide. Mar Drugs 11:2239–2258. doi: 10.3390/md11072239 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B9] 9. Christoffersen K, Lyck S, Winding A. 2002. Microbial activity and bacterial community structure during degradation of microcystins. Aquat Microb Ecol 27:125–136. doi: 10.3354/ame027125 [DOI] [Google Scholar]

[B10] 10. Anderson DM, Hoagland P, Kaoru Y, White AW. 2000. Estimated annual economic impacts from harmful algal blooms (HABs) in the United States. Woods Hole Oceanographic Institution. doi: 10.1575/1912/96 [DOI] [Google Scholar]

[B11] 11. Watson SB, Whitton BA, Higgins SN, Paerl HW, Brooks BW, Wehr JD. 2015. Chapter Ch 20, Harmful algal blooms, p 873–920. In Freshwater algae of North America. Academic Press, Boston. [Google Scholar]

[B12] 12. Huisman J, Codd GA, Paerl HW, Ibelings BW, Verspagen JMH, Visser PM. 2018. Cyanobacterial blooms. Nat Rev Microbiol 16:471–483. doi: 10.1038/s41579-018-0040-1 [DOI] [PubMed] [Google Scholar]

[B13] 13. Paerl HW, Huisman J. 2008. Climate: blooms like it hot. Science 320:57–58. doi: 10.1126/science.1155398 [DOI] [PubMed] [Google Scholar]

[B14] 14. O’Neil JM, Davis TW, Burford MA, Gobler CJ. 2012. The rise of harmful cyanobacteria blooms: the potential roles of eutrophication and climate change. Harmful Algae 14:313–334. doi: 10.1016/j.hal.2011.10.027 [DOI] [Google Scholar]

[B15] 15. Jöhnk KD, Huisman J, Sharples J, Sommeijer B, Visser PM, Stroom JM. 2008. Summer heatwaves promote blooms of harmful cyanobacteria. Glob Chang Biol 14:495–512. doi: 10.1111/j.1365-2486.2007.01510.x [DOI] [Google Scholar]

[B16] 16. Anneville O, Domaizon I, Kerimoglu O, Rimet F, Jacquet S. 2015. “Blue-green algae in a “greenhouse century”? new insights from field data on climate change impacts on cyanobacteria abundance”. Ecosystems 18:441–458. doi: 10.1007/s10021-014-9837-6 [DOI] [Google Scholar]

[B17] 17. Schmidt JR, Wilhelm SW, Boyer GL. 2014. The fate of microcystins in the environment and challenges for monitoring. Toxins (Basel) 6:3354–3387. doi: 10.3390/toxins6123354 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B18] 18. Pacheco ABF, Guedes IA, Azevedo SMFO. 2016. Is qPCR a reliable indicator of cyanotoxin risk in freshwater. Toxins (Basel) 8:172. doi: 10.3390/toxins8060172 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B19] 19. Massey IY, Wu P, Wei J, Luo J, Ding P, Wei H, Yang F. 2020. A mini-review on detection methods of microcystins. Toxins (Basel) 12:641. doi: 10.3390/toxins12100641 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B20] 20. Jones MR, Pinto E, Torres MA, Dörr F, Mazur-Marzec H, Szubert K, Tartaglione L, Dell’Aversano C, Miles CO, Beach DG, McCarron P, Sivonen K, Fewer DP, Jokela J, Janssen EM-L. 2021. CyanoMetDB, a comprehensive public database of secondary metabolites from cyanobacteria. Water Res 196:117017. doi: 10.1016/j.watres.2021.117017 [DOI] [PubMed] [Google Scholar]

[B21] 21. Almuhtaram H, Kibuye FA, Ajjampur S, Glover CM, Hofmann R, Gaget V, Owen C, Wert EC, Zamyadi A. 2021. State of knowledge on early warning tools for cyanobacteria detection. Ecological Indicators 133:108442. doi: 10.1016/j.ecolind.2021.108442 [DOI] [Google Scholar]

[B22] 22. Alivisatos AP, Blaser MJ, Brodie EL, Chun M, Dangl JL, Donohue TJ, Dorrestein PC, Gilbert JA, Green JL, Jansson JK, Knight R, Maxon ME, McFall-Ngai MJ, Miller JF, Pollard KS, Ruby EG, Taha SA, Unified Microbiome Initiative Consortium . 2015. A unified initiative to harness earth’s microbiomes. Science 350:507–508. doi: 10.1126/science.aac8480 [DOI] [PubMed] [Google Scholar]

[B23] 23. Goutman SA, Boss J, Guo K, Alakwaa FM, Patterson A, Kim S, Savelieff MG, Hur J, Feldman EL. 2020. Untargeted metabolomics yields insight into ALS disease mechanisms. J Neurol Neurosurg Psychiatry 91:1329–1338. doi: 10.1136/jnnp-2020-323611 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B24] 24. Zhang H, Wang J, Sun Z, Zurada JM, Pal NR. 2020. Feature selection for neural networks using group lasso regularization. IEEE Trans Knowl Data Eng 32:659–673. doi: 10.1109/TKDE.2019.2893266 [DOI] [Google Scholar]

[B25] 25. Pozzer AC, Gómez PA, Weiss J. 2022. Volatile organic compounds in aquatic ecosystems – detection, origin, significance and applications. Sci Total Environ 838:156155. doi: 10.1016/j.scitotenv.2022.156155 [DOI] [PubMed] [Google Scholar]

[B26] 26. Weisskopf L, Schulz S, Garbeva P. 2021. Microbial volatile organic compounds in intra-kingdom and inter-kingdom interactions. Nat Rev Microbiol 19:391–404. doi: 10.1038/s41579-020-00508-1 [DOI] [PubMed] [Google Scholar]

[B27] 27. Amavizca E, Bashan Y, Ryu C-M, Farag MA, Bebout BM, de-Bashan LE. 2017. Enhanced performance of the Microalga chlorella sorokiniana remotely induced by the plant growth-promoting bacteria Azospirillum brasilense and Bacillus pumilus. Sci Rep 7:41310. doi: 10.1038/srep41310 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B28] 28. Fink P. 2007. Ecological functions of volatile organic compounds in aquatic systems. Mar Freshw Behav Physiol 40:155–168. doi: 10.1080/10236240701602218 [DOI] [Google Scholar]

[B29] 29. Xu Q, Yang L, Yang W, Bai Y, Hou P, Zhao J, Zhou L, Zuo Z. 2017. Volatile organic compounds released from Microcystis flos-aquae under nitrogen sources and their toxic effects on Chlorella Vulgaris. Ecotoxicology and Environmental Safety 135:191–200. doi: 10.1016/j.ecoenv.2016.09.027 [DOI] [PubMed] [Google Scholar]

[B30] 30. Zuo Z-J, Zhu Y-R, Bai Y-L, Wang Y. 2012. Volatile communication between Chlamydomonas reinhardtii cells under salt stress. Biochemical Systematics and Ecology 40:19–24. doi: 10.1016/j.bse.2011.09.007 [DOI] [Google Scholar]

[B31] 31. Schulz-Bohm K, Geisen S, Wubs ERJ, Song C, de Boer W, Garbeva P. 2017. The prey’s scent – volatile organic compound mediated interactions between soil bacteria and their protist predators. ISME J 11:817–820. doi: 10.1038/ismej.2016.144 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B32] 32. Halsey KH, Giovannoni SJ, Graus M, Zhao Y, Landry Z, Thrash JC, Vergin KL, de Gouw J. 2017. Biological cycling of volatile organic carbon by phytoplankton and bacterioplankton. Limnol Oceanogr 62:2650–2661. doi: 10.1002/lno.10596 [DOI] [Google Scholar]

[B33] 33. Moore ER, Weaver AJ, Davis EW, Giovannoni SJ, Halsey KH. 2022. Metabolism of key atmospheric volatile organic compounds by the marine heterotrophic bacterium Pelagibacter HTCC1062 (SAR11). Environ Microbiol 24:212–222. doi: 10.1111/1462-2920.15837 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B34] 34. Aneja VP, Schlesinger WH, Erisman JW. 2009. Effects of agriculture upon the air quality and climate: research, policy, and regulations. Environ Sci Technol 43:4234–4240. doi: 10.1021/es8024403 [DOI] [PubMed] [Google Scholar]

[B35] 35. Ozaki K, Ohta A, Iwata C, Horikawa A, Tsuji K, Ito E, Ikai Y, Harada K-I. 2008. Lysis of cyanobacteria with volatile organic compounds. Chemosphere 71:1531–1538. doi: 10.1016/j.chemosphere.2007.11.052 [DOI] [PubMed] [Google Scholar]

[B36] 36. Ye C, Yang Y, Xu Q, Ying B, Zhang M, Gao B, Ni B, Yakefu Z, Bai Y, Zuo Z. 2018. Volatile organic compound emissions from Microcystis aeruginosa under different phosphorus sources and concentrations. Phycological Res 66:15–22. doi: 10.1111/pre.12201 [DOI] [Google Scholar]

[B37] 37. Zuo Z, Yang Y, Xu Q, Yang W, Zhao J, Zhou L. 2018. Effects of phosphorus sources on volatile organic compound emissions from Microcystis flos-aquae and their toxic effects on Chlamydomonas reinhardtii. Environ Geochem Health 40:1283–1298. doi: 10.1007/s10653-017-0055-y [DOI] [PubMed] [Google Scholar]

[B38] 38. Ikawa M, Sasner JJ, Haney JF. 2001. Activity of cyanobacterial and algal odor compounds found in Lake waters on green alga Chlorella pyrenoidosa growth. Hydrobiologia 443:19–22. doi: 10.1023/A:1017535801766 [DOI] [Google Scholar]

[B39] 39. Yamashita R, Bober B, Kanei K, Arii S, Tsuji K, Harada K-I. 2020. Analytical technique optimization on the detection of β-cyclocitral in Microcystis species. Molecules 25:832. doi: 10.3390/molecules25040832 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B40] 40. Zuo Z, Zhu Y, Bai Y, Wang Y. 2012. Acetic acid-induced programmed cell death and release of volatile organic compounds in Chlamydomonas reinhardtii. Plant Physiol Biochem 51:175–184. doi: 10.1016/j.plaphy.2011.11.003 [DOI] [PubMed] [Google Scholar]

[B41] 41. Liu M, Wu T, Zhao X, Zan F, Yang G, Miao Y. 2021. Cyanobacteria blooms potentially enhance volatile organic compound (VOC) emissions from a Eutrophic Lake: field and experimental evidence. Environ Res 202:111664. doi: 10.1016/j.envres.2021.111664 [DOI] [PubMed] [Google Scholar]

[B42] 42. Zuo Z, Yang L, Chen S, Ye C, Han Y, Wang S, Ma Y. 2018. Effects of nitrogen nutrients on the volatile organic compound emissions from Microcystis aeruginosa. Ecotoxicol Environ Saf 161:214–220. doi: 10.1016/j.ecoenv.2018.05.095 [DOI] [PubMed] [Google Scholar]

[B43] 43. Bižić M, Klintzsch T, Ionescu D, Hindiyeh MY, Günthel M, Muro-Pastor AM, Eckert W, Urich T, Keppler F, Grossart H-P. 2020. Aquatic and terrestrial cyanobacteria produce methane. Sci Adv 6:eaax5343. doi: 10.1126/sciadv.aax5343 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B44] 44. Santos AB, Fernandes AS, Wagner R, Jacob-Lopes E, Zepka LQ. 2016. Biogeneration of volatile organic compounds produced by Phormidium autumnale in heterotrophic bioreactor. J Appl Phycol 28:1561–1570. doi: 10.1007/s10811-015-0740-0 [DOI] [Google Scholar]

[B45] 45. Lindberg P, Park S, Melis A. 2010. Engineering a platform for photosynthetic isoprene production in cyanobacteria, using Synechocystis as the model organism. Metab Eng 12:70–79. doi: 10.1016/j.ymben.2009.10.001 [DOI] [PubMed] [Google Scholar]

[B46] 46. O’Dowd CD, de Leeuw G. 2007. Marine aerosol production: a review of the current knowledge. Philos Trans A Math Phys Eng Sci 365:1753–1774. doi: 10.1098/rsta.2007.2043 [DOI] [PubMed] [Google Scholar]

[B47] 47. Moore ER, Davie-Martin CL, Giovannoni SJ, Halsey KH. 2020. Pelagibacter metabolism of diatom‐derived volatile organic compounds imposes an energetic tax on photosynthetic carbon fixation. Environ Microbiol 22:1720–1733. doi: 10.1111/1462-2920.14861 [DOI] [PubMed] [Google Scholar]

[B48] 48. Kieber DJ, McDaniel J, Mopper K. 1989. Photochemical source of biological substrates in sea water: implications for carbon cycling. Nature 341:637–639. doi: 10.1038/341637a0 [DOI] [Google Scholar]

[B49] 49. Vuorio E, Thiel K, Fitzpatrick D, Huokko T, Kämäräinen J, Dandapani H, Aro E-M, Kallio P. 2021. Hydrocarbon desaturation in cyanobacterial thylakoid membranes is linked with acclimation to suboptimal growth temperatures. Front Microbiol 12:781864. doi: 10.3389/fmicb.2021.781864 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B50] 50. Lea-Smith DJ, Ortiz-Suarez ML, Lenn T, Nürnberg DJ, Baers LL, Davey MP, Parolini L, Huber RG, Cotton CAR, Mastroianni G, Bombelli P, Ungerer P, Stevens TJ, Smith AG, Bond PJ, Mullineaux CW, Howe CJ. 2016. Hydrocarbons are essential for optimal cell size, division, and growth of cyanobacteria. Plant Physiol 172:1928–1940. doi: 10.1104/pp.16.01205 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B51] 51. Qiao Y, Wang W, Lu X. 2020. High light induced alka(e)ne biodegradation for lipid and redox homeostasis in cyanobacteria. Front. Microbiol 11:1659. doi: 10.3389/fmicb.2020.01659 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B52] 52. Berla BM, Saha R, Maranas CD, Pakrasi HB. 2015. Cyanobacterial alkanes modulate photosynthetic cyclic electron flow to assist growth under cold stress. Sci Rep 5:14894. doi: 10.1038/srep14894 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B53] 53. Oregon Health Authority. Accession May 15, 2023. https://www.oregon.gov/oha/ph/healthyenvironments/recreation/harmfulalgaeblooms/pages/blue-greenalgaeadvisories.aspx. [Google Scholar]

[B54] 54. United States Environmental Protection Agency. Accession May 15, 2023. https://www.epa.gov/cyanohabs/epa-drinking-water-health-advisories-cyanotoxins. [Google Scholar]

[B55] 55.Agency U.S. Environmental ProtectionHealth and ecological criteria division. Washington, DC. 2019. Recommended human health recreational ambient water quality criteria or swimming advisories for microcystins and cylindrospermopsin. EPA Document Number: 822-R-19-001.

[B56] 56. Carmichael WW, Drapeau C, Anderson DM. 2000. Harvesting of Aphanizomenon flos-aquae ralfs ex born. J Appl Phycol 12:585–595. doi: 10.1023/A:1026506713560 [DOI] [Google Scholar]

[B57] 57. Gugger M, Lyra C, Henriksen P, Couté A, Humbert J-F, Sivonen K. 2002. Phylogenetic comparison of the cyanobacterial genera Anabaena and Aphanizomenon. Int J Syst Evol Microbiol 52:1867–1880. doi: 10.1099/00207713-52-5-1867 [DOI] [PubMed] [Google Scholar]

[B58] 58. Dreher TW, Davis EW, Mueller RS. 2021. Complete genomes derived by directly sequencing freshwater bloom populations emphasize the significance of the genus level ADA clade within the Nostocales. Harmful Algae 103:102005. doi: 10.1016/j.hal.2021.102005 [DOI] [PubMed] [Google Scholar]

[B59] 59. Rousso BZ, Bertone E, Stewart R, Hamilton DP. 2020. A systematic literature review of forecasting and predictive models for cyanobacteria blooms in freshwater lakes. Water Res 182:115959. doi: 10.1016/j.watres.2020.115959 [DOI] [PubMed] [Google Scholar]

[B60] 60. Madsen R, Lundstedt T, Trygg J. 2010. Chemometrics in metabolomics—a review in human disease diagnosis. Anal Chim Acta 659:23–33. doi: 10.1016/j.aca.2009.11.042 [DOI] [PubMed] [Google Scholar]

[B61] 61. Mallick H, Franzosa EA, Mclver LJ, Banerjee S, Sirota-Madi A, Kostic AD, Clish CB, Vlamakis H, Xavier RJ, Huttenhower C. 2019. Predictive metabolomic profiling of microbial communities using amplicon or metagenomic sequences. Nat Commun 10:3136. doi: 10.1038/s41467-019-10927-1 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B62] 62. Morrow AL, Lagomarcino AJ, Schibler KR, Taft DH, Yu Z, Wang B, Altaye M, Wagner M, Gevers D, Ward DV, Kennedy MA, Huttenhower C, Newburg DS. 2013. Early microbial and metabolomic signatures predict later onset of necrotizing enterocolitis in preterm infants. Microbiome 1:13. doi: 10.1186/2049-2618-1-13 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B63] 63. Wilson RM, Tfaily MM, Kolton M, Johnston ER, Petro C, Zalman CA, Hanson PJ, Heyman HM, Kyle JE, Hoyt DW, Eder EK, Purvine SO, Kolka RK, Sebestyen SD, Griffiths NA, Schadt CW, Keller JK, Bridgham SD, Chanton JP, Kostka JE. 2021. Soil metabolome response to whole-ecosystem warming at the spruce and peatland responses under changing environments experiment. Proc Natl Acad Sci U S A 118:e2004192118. doi: 10.1073/pnas.2004192118 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B64] 64. Lawson CA, Possell M, Seymour JR, Raina J-B, Suggett DJ. 2019. Coral endosymbionts (Symbiodiniaceae) emit species-specific volatilomes that shift when exposed to thermal stress. Sci Rep 9:17395. doi: 10.1038/s41598-019-53552-0 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B65] 65. Meskhidze N, Sabolis A, Reed R, Kamykowski D. 2015. Quantifying environmental stress-induced emissions of algal isoprene and monoterpenes using laboratory measurements. Biogeosciences 12:637–651. doi: 10.5194/bg-12-637-2015 [DOI] [Google Scholar]

[B66] 66. Dani KGS, Loreto F. 2017. Trade-off between dimethyl sulfide and isoprene emissions from marine phytoplankton. Trends Plant Sci 22:361–372. doi: 10.1016/j.tplants.2017.01.006 [DOI] [PubMed] [Google Scholar]

[B67] 67. Noecker C, Eng A, Srinivasan S, Theriot CM, Young VB, Jansson JK, Fredricks DN, Borenstein E, Sanchez LM. 2016. Metabolic model-based integration of microbiome taxonomic and metabolomic profiles elucidates mechanistic links between ecological and metabolic variation. mSystems 1:e00013-15. doi: 10.1128/mSystems.00013-15 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B68] 68. Morton JT, Aksenov AA, Nothias LF, Foulds JR, Quinn RA, Badri MH, Swenson TL, Van Goethem MW, Northen TR, Vazquez-Baeza Y, Wang M, Bokulich NA, Watters A, Song SJ, Bonneau R, Dorrestein PC, Knight R. 2019. Learning representations of microbe–metabolite interactions. Nat Methods 16:1306–1314. doi: 10.1038/s41592-019-0616-3 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B69] 69. Quinn TP, Erb I. 2021. Examining microbe–metabolite correlations by linear methods. Nat Methods 18:37–39. doi: 10.1038/s41592-020-01006-1 [DOI] [PubMed] [Google Scholar]

[B70] 70. Morton JT, McDonald D, Aksenov AA, Nothias LF, Foulds JR, Quinn RA, Badri MH, Swenson TL, Van Goethem MW, Northen TR, Vazquez-Baeza Y, Wang M, Bokulich NA, Watters A, Song SJ, Bonneau R, Dorrestein PC, Knight R. 2021. Reply to: examining microbe–metabolite correlations by linear methods. Nat Methods 18:40–41. doi: 10.1038/s41592-020-01007-0 [DOI] [PubMed] [Google Scholar]

[B71] 71. Agger SA, Lopez-Gallego F, Hoye TR, Schmidt-Dannert C. 2008. Identification of sesquiterpene synthases from Nostoc punctiforme PCC 73102 and Nostoc sp. strain PCC 7120. J Bacteriol 190:6084–6096. doi: 10.1128/JB.00759-08 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B72] 72. Merel S, Walker D, Chicana R, Snyder S, Baurès E, Thomas O. 2013. State of knowledge and concerns on cyanobacterial blooms and cyanotoxins. Environ Int 59:303–327. doi: 10.1016/j.envint.2013.06.013 [DOI] [PubMed] [Google Scholar]

[B73] 73. Muñoz-González C, Canon F, Feron G, Guichard E, Pozo-Bayón M. 2019. Assessment wine aroma persistence by using an in vivo PTR-ToF-MS approach and its relationship with salivary parameters. Molecules 24:1277. doi: 10.3390/molecules24071277 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B74] 74. Zhang K, Lin TF, Zhang T, Li C, Gao N. 2013. Characterization of typical taste and odor compounds formed by Microcystis aeruginosa. J Environ Sci (China) 25:1539–1548. doi: 10.1016/s1001-0742(12)60232-0 [DOI] [PubMed] [Google Scholar]

[B75] 75. Shao J, Peng L, Luo S, Yu G, Gu J, Lin S, Li R. 2013. First report on the allelopathic effect of Tychonema bourrellyi (cyanobacteria) against Microcystis aeruginosa (cyanobacteria). J Appl Phycol 25:1567–1573. doi: 10.1007/s10811-012-9969-z [DOI] [Google Scholar]

[B76] 76. Liu J, Yu Q, Ye B, Zhu K, Yin J, Zheng T, Xu S, Sun Q, Li Y, Zuo Z. 2021. Programmed cell death of Chlamydomonas reinhardtii induced by three cyanobacterial volatiles Β-ionone, limonene and longifolene. Sci Total Environ 762:144539. doi: 10.1016/j.scitotenv.2020.144539 [DOI] [PubMed] [Google Scholar]

[B77] 77. Shao J, Xu Y, Wang Z, Jiang Y, Yu G, Peng X, Li R. 2011. Elucidating the toxicity targets of Β-ionone on photosynthetic system of Microcystis aeruginosa NIES-843 (cyanobacteria). Aquat Toxicol 104:48–55. doi: 10.1016/j.aquatox.2011.03.014 [DOI] [PubMed] [Google Scholar]

[B78] 78. Walsh K, Jones GJ, Dunstan RH. 1998. Effect of high irradiance and iron on volatile odour compounds in the cyanobacterium Microcystis aeruginosa. Phytochemistry 49:1227–1239. doi: 10.1016/s0031-9422(97)00943-6 [DOI] [PubMed] [Google Scholar]

[B79] 79. Suurnäkki S, Gomez-Saez GV, Rantala-Ylinen A, Jokela J, Fewer DP, Sivonen K. 2015. Identification of geosmin and 2-methylisoborneol in cyanobacteria and molecular detection methods for the producers of these compounds. Water Res 68:56–66. doi: 10.1016/j.watres.2014.09.037 [DOI] [PubMed] [Google Scholar]

[B80] 80. Hellweger FL, Schampera C, Martin RM, Eigemann F, Smith DJ, Dick GJ, Wilhelm SW. 2022. Models predict planned phosphorus load reduction will make lake erie more toxic. Science 378:eade2277. doi: 10.1126/science.ade2277 [DOI] [PubMed] [Google Scholar]

[B81] 81. Höckelmann C, Jüttner F. 2004. Volatile organic compound (VOC) analysis and sources of limonene, cyclohexanone and straight chain aldehydes in axenic cultures of Calothrix and Plectonema. Water Sci Technol 49:47–54. [PubMed] [Google Scholar]

[B82] 82. Yang R, Zhu L, Li T, Zhu L, Ye Z, Zhang D. 2021. Photosynthetic conversion of CO2 into pinene using engineered Synechococcus sp. Front Bioeng Biotechnol 9. doi: 10.3389/fbioe.2021.779437 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B83] 83. Pattanaik B, Lindberg P. 2015. Terpenoids and their biosynthesis in cyanobacteria. Life (Basel) 5:269–293. doi: 10.3390/life5010269 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B84] 84. Zhao J, Yang L, Zhou L, Bai Y, Wang B, Hou P, Xu Q, Yang W, Zuo Z. 2016. Inhibitory effects of Eucalyptol and limonene on the photosynthetic abilities in Chlorella vulgaris (Chlorophyceae). Phycologia 55:696–702. doi: 10.2216/16-38.1 [DOI] [Google Scholar]

[B85] 85. Hu X, Liu Y, Zeng G, Hu X, Wang Y, Zeng X. 2014. Effects of limonene stress on the growth of and microcystin release by the freshwater cyanobacterium Microcystis aeruginosa FACHB-905. Ecotoxicol Environ Saf 105:121–127. doi: 10.1016/j.ecoenv.2014.01.023 [DOI] [PubMed] [Google Scholar]

[B86] 86. Ghanbari J, Khajoei-Nejad G, Erasmus SW, van Ruth SM. 2019. Identification and characterisation of volatile fingerprints of saffron stigmas and petals using PTR-TOF-MS: influence of nutritional treatments and corm provenance. Ind Crops and Prod 141:111803. doi: 10.1016/j.indcrop.2019.111803 [DOI] [Google Scholar]

[B87] 87. Jalliffier-Merlon E, Marty J-C, Denant V, Saliot A. 1991. Phytoplanktonic sources of volatile aldehydes in the river Rhône estuary. Estuarine Coastal and Shelf Science 32:463–482. doi: 10.1016/0272-7714(91)90035-A [DOI] [Google Scholar]

[B88] 88. Coates RC, Podell S, Korobeynikov A, Lapidus A, Pevzner P, Sherman DH, Allen EE, Gerwick L, Gerwick WH. 2014. Characterization of cyanobacterial hydrocarbon composition and distribution of biosynthetic pathways. PLoS One 9:e85140. doi: 10.1371/journal.pone.0085140 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B89] 89. Parveen H, Yazdani SS. 2022. Insights into cyanobacterial alkane biosynthesis. J Ind Microbiol Biotechnol 49:kuab075. doi: 10.1093/jimb/kuab075 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B90] 90. Xue G, Fiedler AT, Martinho M, Münck E, Que L. 2008. Insights into the P-to-Q conversion in the catalytic cycle of methane monooxygenase from a synthetic model system. Proc Natl Acad Sci USA 105:20615–20620. doi: 10.1073/pnas.0808512105 [DOI] [Google Scholar]

[B91] 91. Mitschke J, Vioque A, Haas F, Hess WR, Muro-Pastor AM. 2011. Dynamics of transcriptional start site selection during nitrogen stress-induced cell differentiation in Anabaena sp. PCC7120. Proc Natl Acad Sci U S A 108:20130–20135. doi: 10.1073/pnas.1112724108 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B92] 92. Hernández-Prieto MA, Semeniuk TA, Giner-Lamia J, Futschik ME. 2016. The transcriptional landscape of the photosynthetic model cyanobacterium Synechocystis sp. PCC6803. Sci Rep 6:22168. doi: 10.1038/srep22168 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B93] 93. Dick GJ, Duhaime MB, Evans JT, Errera RM, Godwin CM, Kharbush JJ, Nitschky HS, Powers MA, Vanderploeg HA, Schmidt KC, Smith DJ, Yancey CE, Zwiers CC, Denef VJ. 2021. The genetic and ecophysiological diversity of Microcystis. Environ Microbiol 23:7278–7313. doi: 10.1111/1462-2920.15615 [DOI] [PubMed] [Google Scholar]

[B94] 94. Zilliges Y, Kehr J-C, Meissner S, Ishida K, Mikkat S, Hagemann M, Kaplan A, Börner T, Dittmann E. 2011. The cyanobacterial hepatotoxin microcystin binds to proteins and increases the fitness of Microcystis under oxidative stress conditions. PLoS One 6:e17615. doi: 10.1371/journal.pone.0017615 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B95] 95. Dziallas C, Grossart H-P. 2011. Increasing oxygen radicals and water temperature select for toxic Microcystis sp. PLoS One 6:e25569. doi: 10.1371/journal.pone.0025569 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B96] 96. Díez-Quijada L, Prieto AI, Guzmán-Guillén R, Jos A, Cameán AM. 2019. Occurrence and toxicity of microcystin congeners other than MC-LR and MC-RR: a review. Food Chem Toxicol 125:106–132. doi: 10.1016/j.fct.2018.12.042 [DOI] [PubMed] [Google Scholar]

[B97] 97. Van de Waal DB, Verspagen JMH, Lürling M, Van Donk E, Visser PM, Huisman J. 2009. The ecological stoichiometry of toxins produced by harmful cyanobacteria: an experimental test of the carbon-nutrient balance hypothesis. Ecol Lett 12:1326–1335. doi: 10.1111/j.1461-0248.2009.01383.x [DOI] [PubMed] [Google Scholar]

[B98] 98. Sparks DL, Page AL, Helmke PA, Loeppert RH. (ed) 2020. Methods of Soil Anaylsis, Part 3: Chemical Methods.Wiley.

[B99] 99. Halsey KH, Milligan AJ, Behrenfeld MJ. 2010. Physiological optimization underlies growth rate-independent chlorophyll-specific gross and net primary production. Photosynth Res 103:125–137. doi: 10.1007/s11120-009-9526-z [DOI] [PubMed] [Google Scholar]

[B100] 100. Ritchie RJ. 2006. Consistent sets of spectrophotometric chlorophyll equations for acetone, methanol and ethanol solvents. Photosynth Res 89:27–41. doi: 10.1007/s11120-006-9065-9 [DOI] [PubMed] [Google Scholar]

[B101] 101. Holzinger R. 2015. PTRwid: a new widget tool for processing PTR-TOF-MS data. Atmos Meas Tech 8:3903–3922. doi: 10.5194/amt-8-3903-2015 [DOI] [Google Scholar]

[B102] 102. Love MI, Huber W, Anders S. 2014. Moderated estimation of fold change and dispersion for RNA-Seq data with DESeq2. Genome Biol 15:550. doi: 10.1186/s13059-014-0550-8 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B103] 103. Yáñez-Serrano AM, Filella I, LLusià J, Gargallo-Garriga A, Granda V, Bourtsoukidis E, Williams J, Seco R, Cappellin L, Werner C, de Gouw J, Peñuelas J. 2021. GLOVOCS - master compound assignment guide for proton transfer reaction mass spectrometry users. Atmospheric Environment 244:117929. doi: 10.1016/j.atmosenv.2020.117929 [DOI] [Google Scholar]

[B104] 104. Friedman JH, Hastie T, Tibshirani R. 2010. Regularization paths for generalized linear regression models via coordinate descent. J. Stat. Soft 33:1–22. doi: 10.18637/jss.v033.i01 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B105] 105. Kenneth Tay J, Narashimhan B, Hastie T. 2023. Elastic net regularization paths for all generalized linear models. J Stat Soft 108:1–31. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B106] 106. Bolaños LM, Karp-Boss L, Choi CJ, Worden AZ, Graff JR, Haëntjens N, Chase AP, Della Penna A, Gaube P, Morison F, Menden-Deuer S, Westberry TK, O’Malley RT, Boss E, Behrenfeld MJ, Giovannoni SJ. 2020. Small phytoplankton dominate western North Atlantic biomass. ISME J 14:1663–1674. doi: 10.1038/s41396-020-0636-0 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B107] 107. Roush D, Giraldo-Silva A, Garcia-Pichel F. 2021. Cydrasil 3, a curated 16S rRNA gene reference package and web app for cyanobacterial phylogenetic placement. Sci Data 8:230. doi: 10.1038/s41597-021-01015-5 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B108] 108. Letunic I, Bork P. 2021. Interactive tree of life (iTOL) v5: an online tool for phylogenetic tree display and annotation. Nucleic Acids Res 49:W293–W296. doi: 10.1093/nar/gkab301 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B109] 109. McMurdie PJ, Holmes S. 2013. phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS One 8:e61217. doi: 10.1371/journal.pone.0061217 [DOI] [PMC free article] [PubMed] [Google Scholar]

[B110] 110. Gu Z, Eils R, Schlesner M. 2016. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics 32:2847–2849. doi: 10.1093/bioinformatics/btw313 [DOI] [PubMed] [Google Scholar]

[B111] 111. Blighe K, Rana S, Lewis M. 2021. Enhancedvolcano: publication-ready volcano plots with enhanced colouring and labeling. Available from: https://github.com/kevinblighe/EnhancedVolcano

PERMALINK

The volatilome reveals microcystin concentration, microbial composition, and oxidative stress in a critical Oregon freshwater lake

Lindsay Collart

Duo Jiang

Kimberly H Halsey

Roles

ABSTRACT

IMPORTANCE

INTRODUCTION

RESULTS AND DISCUSSION

Upper Klamath Lake chemical and microbial composition

Fig 1.

TABLE 1.

Fig 2.

Fig 3.

Fig 4.

Microcystin toxin prediction using the volatilome

TABLE 2.

TABLE 3.

Fig 5.

Predicting microbial community composition using the volatilome

Fig 6.

Fig 7.

Conclusion

MATERIALS AND METHODS

Water sample collection

Environmental parameters

Detection of VOCs

VOC data processing

VOC-based elastic net models predicting microcystin concentration

“Low cost” and “low + high-cost” regression models predicting microcystin concentration

DNA extraction and sequencing

16S rRNA gene amplicon analysis

VOC-based models of microbial community composition

ACKNOWLEDGMENTS

Contributor Information

DATA AVAILABILITY

SUPPLEMENTAL MATERIAL

REFERENCES

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases