Seasonal Variability, 2009


Analysis of median concentrations of total microcystins, chlorophyll a, total nutrients, and dissolved nutrients during the 2009 sample season reveals potentially important seasonal patterns among these parameters (fig. 6). Only data collected in 2009 are shown because total microcystin concentrations measured in 2008 were significantly lower than in 2007 or 2009 and because concentrations measured in 2007 do not included the dissolved fraction. Correlation analysis was performed between total microcystins or microcystins in the dissolved or large (> 63 μm) particulate fractions measured in 2009 and the suite of environmental variables used for the interannual correlation analysis (table 5).


Chlorophyll a


In 2009, the first lakewide maxima in chlorophyll a concentrations occurred while total microcystin concentrations were lowest (fig. 6A). Total microcystin concentrations began to increase in mid-July near the end of the A. flos-aquae-dominated bloom decline, then increased more rapidly during the bloom recovery, beginning July 21 and 27. A corresponding increase in M. aeruginosa cell density also was recorded at sites near those used in the current study (J. Kann, Aquatic Ecosystem Sciences, LLC, unpub. data, 2011). Between August and mid-September, total microcystin concentrations followed a similar pattern to chlorophyll a concentrations, peaking on August 17 and declining thereafter until the end of the season. As with results of the interannual correlation analysis, no correlation was found between chlorophyll a and large particulate (cell associated) microcystin concentrations in 2009 (table 5), which supports the low detection of microcystin concentrations during the first bloom and the observed high concentrations during the second bloom.


Total and Dissolved Nutrients 


As seen in previous years in Upper Klamath Lake, total nitrogen and total phosphorus increased between May and October 2009 (fig. 6B). At the onset of both A. flos-aquae-dominated blooms (during late-June and late-August; fig. 6A), total nitrogen increased rapidly with chlorophyll a, suggesting that major bloom expansions during these periods primarily were associated with atmospheric nitrogen sequestration by nitrogen fixation. Benthic fluxes also add nitrogen to the water column (Kuwabara and others, 2007), as do riverine inputs, but these sources are smaller by comparison (Kann and Walker, 1999). Total phosphorus increased more steadily than did total nitrogen during the 2009 season (fig. 6B), which suggests that internal loading from lake sediments, the most important source of phosphorus to this system, is less of an episodic phenomenon. Results of weekly sampling in 2009 also show that microcystin concentrations were highest after the mid-season increase in total phosphorus and nitrogen following the major bloom decline, and that, although the concentration of dissolved inorganic nitrogen decreased after July 27, concentrations were still sufficient to promote growth of non-diazotrophic, toxigenic strains of M. aeruginosa. Total phosphorus correlated positively with total microcystins (r = 0.61, p = 0.02), dissolved microcystins (r = 0.69, p = 0.01) and with microcystins in the large particulate fraction (r = 0.63, p = 0.01; table 5), and no correlations were found between microcystins and total nitrogen. This suggests a possible relation between microcystin and total phosphorus concentrations in 2009. It is not possible with the available data to determine whether the presence of phosphorus directly influences microcystin occurrence or if both factors are related to some other common variable, particularly because the correlation between large particulate microcystins and total phosphorus in all years combined was not significant at p < 0.05. However, it can be hypothesized that the positive correlation with total phosphorus concentrations in 2009 is due to an indirect association between microcystin occurrence (or M. aeruginosa growth) and phosphorus availability, because bioavailable phosphorus appears to regulate the A. flos-aquae-dominated bloom and given the apparent dependence of M. aeruginosa growth on nitrogen-fixation by A. flos-aquae. 


Coincident with minimum concentrations of chlorophyll a near the end of July, concentrations of dissolved inorganic nitrogen (DIN) and dissolved inorganic phosphorus (DIP) reached a seasonal maximum as a result of nutrient release during cell lysis and decomposition of senesced phytoplankton (figs. 6C and 6D). These sharp increases and peaks in DIN and DIP concentrations occurred 2 weeks prior to the start of the increase in total microcystin concentrations on July 20 and before the peak in total microcystin concentrations was observed on August 10 (fig. 2). This suggests that the increase in available nitrogen and phosphorus promoted growth of both A. flos-aquae (as indicated by increasing chlorophyll a) and toxigenic M. aeruginosa (as indicated by increased microcystin concentrations). Diazotrophic species, such as A. flos-aquae, use DIN preferentially as a nitrogen source over nitrogen-fixation, which is a far more energy-consuming process, so it is likely that these organisms use N₂ after DIN is depleted. Dissolved microcystins correlated positively with DIP (r = 0.53, p = 0.05) and negatively with DIN:DIP ratio (r = -0.55, p = 0.04), but, as with total nitrogen, no correlations were observed between DIN and microcystins in any size fraction.


Nutrient Ratios


The late-July 2009 increase in total microcystin concentrations coincided with low TN:TP ratios and the second A. flos-aquae­-dominated bloom; TN:TP ratios were higher during the first bloom than when this second bloom occurred (figs. 6A and 7A). Concentrations of total nitrogen and total phosphorus increased over the season (fig. 6B), so the decrease in TN:TP between the first and second blooms was due to a greater increase in total phosphorus, relative to total nitrogen concentrations, possibly as a consequence of decreased nitrogen-fixation during the A. flos-aquae-bloom decline. TN:TP ratios correlated negatively with total, dissolved, and particulate (cell associated) microcystin concentrations (total microcystins: r = -0.69, p = 0.01; dissolved microcystins: r = -0.73, p = < 0.01; large particulate microcystins: r = -0.69, p = 0.01); in correlation analysis using data from all years, only large particulate microcystins were significantly correlated. This result is consistent with other studies (Jacoby and others, 2000; Kotak and others, 2000; Xie and others, 2003), but does not support interspecific competition for nitrogen as the primary reason for diazotrophic A. flos-aquae dominance (by biovolume) over non-diazotrophic M. aeruginosa in Upper Klamath Lake, at least during mid- to late-summer.


Microcystin concentrations began to increase during the period of higher DIN:DIP ratios (prior to August 1; fig. 7B) that resulted from the release of nutrients during the first major bloom decline. Concentrations of dissolved microcystins initially peaked along with DIN:DIP (after August 1), but the peak in particulate fraction microcystins was not observed until after the sharp decrease in DIN:DIP, which occurred as nutrient uptake increased during the second A. flos-aquae-dominated bloom (figs. 2 and 7B). Therefore, it appears that intracellular toxin production was not adversely affected by the low DIN:DIP ratios (and the accompanying low ammonia concentrations, fig. 6D) after August 10. Total, dissolved, and large particulate (cell associated) microcystins were negatively correlated with DIN:DIP ratios (total microcystins: r = 0.70, p < 0.01; large particulate microcystins: r = -0.71, p < 0.01; dissolved microcystins: r = -0.55, p = 0.04), in agreement with results of the interannual correlation analysis and with the observed decrease in DIN:DIP prior to the peak in particulate microcystin concentrations. 


Particulate Carbon, Nitrogen, and Phosphorus


Changes in median concentrations of total particulate carbon (TPC), total particulate nitrogen (TPN), and total particulate phosphorus (TPP), which represent changes in the population density of primarily A. flos-aquae over time in Upper Klamath Lake, followed patterns similar to that of chlorophyll a (figs. 6A and 8) in 2009, increasing during the first bloom in late June-early July, decreasing during the bloom decline, and increasing again, along with total microcystins, as the second bloom developed in mid-August. As with chlorophyll a, the relatively low median particulate nutrient concentrations in the last half of July corresponded withpeaks in median concentrations of dissolved nutrients (figs. 6C and 6D), in that particulate nutrients transitioned into dissolved nutrients during the bloom decline. However, no correlation was found between total particulate nutrient and microcystin concentrations (table 5).


Median ratios of TPN:TPP decreased between early July and early August similarly to the pattern observed for the ratio of total nitrogen to total phosphorus (figs. 7A and 8D). Median TPN:TPP ratios of 24–25 were observed along with peak chlorophyll a concentrations during the first A. flos-aquae bloom between June 16 and 23, whereas the median TPN:TPP ratio was closer to 14 during the second peak in the A. flos-aquae bloom on August 17. Concentrations of particulate inorganic phosphorus (PIP) were 29 ±12 percent of TPP and exhibited no clear seasonal trend (data not shown), which indicates that ratios of organic PN:PP were closer to 34 or 35 during the first bloom and near 15 during the second bloom. Although the organic particulate material was not composed exclusively of A. flos-aquae, the dominance of A. flos-aquae in phytoplankton biomass of Upper Klamath Lake indicates a lower overall N:P ratio in cells comprising the second A. flos-aquae-dominated bloom, which is consistent with a switch from nutrient-limited to nutrient-replete conditions (Klausmeier and others, 2004). The most rapid increase and seasonal maxima of total microcystin concentrations (and M. aeruginosa cell densities at adjacent sample sites; J. Kann, Aquatic Ecosystem Sciences, LLC, unpub. data, 2011) occurred with the lowest seasonal TPN:TPP ratios, which accounts for the significant correlation between total particulate nutrient ratios and total and large particulate (cell associated) microcystin concentrations.


Continuous Monitor and Meteorological Variables


In 2009, total and dissolved microcystin concentrations were positively correlated with water column stability measured at site MDN (total: r = 0.51, p = 0.05; dissolved: r = 0.61, p = 0.02; the p-value was low but not significant at p < 0.05 between large particulate microcystins and water column stability: r = 0.46, p = 0.09). Total microcystin concentrations also correlated negatively near significance (p < 0.1) with daily median wind speed (r = -0.48, p = 0.08), which, together with the positive correlation between microcystin concentrations and water column stability, agree with results of the interannual correlation analysis that shows accumulation of toxigenic M. aeruginosa cells, as with A. flos-aquae colonies, occurs in a stable water column. Elevated pH and dissolved oxygen concentrations are characteristic of dense phytoplankton growth, and the non-significant correlation observed between water column pH, dissolved oxygen concentrations, and total or particulate fraction microcystin concentrations may indicate that the increased productivity during bloom development and the large decreases in pH and dissolved oxygen concentrations between the first and second bloom periods (during bloom decline) do not influence microcystin occurrence in the lake.


First posted May 30, 2012