2.4.1. Description and Significance
Phytoplankton (microscopic algae), including cyanobacteria (also called blue-green algae) photosynthesize and serve as the base of the food chain in lakes, rivers, streams, estuaries, and oceans. Some species thrive in salt water, some in fresh water, and some tolerate wide ranges of salinity. Under certain conditions of nutrients, light, salinity and flow, these organisms can propagate rapidly and result in very high concentrations of the algae, creating what is called a “bloom”, which can have significant impacts on the local ecology of a river or lake (Figure 2.18).
Algal blooms are often described as nuisances because of the odor and unsightliness of algal scum and the green water that often accompanies them. However, the potential impacts go well beyond being a nuisance. Blooms, in addition to being clearly visible events, often induce high oxygen production during the daylight hours (due to photosynthesis), followed at night by very low oxygen levels (due to respiration). Other effects occur when blooms are so dense that sunlight cannot reach the native submerged aquatic vegetation, reducing the plants’ ability to photosynthesize and grow. Also, when the bloom biomass decays, dissolved oxygen levels are decreased. As a consequence, survival of fish and other aquatic organisms may become threatened by low oxygen and reduced food and habitat caused by algal blooms.
Some cyanobacteria species produce toxins (cyanotoxins) that can reach high levels in a bloom, potentially creating public health problems and causing widespread deaths of fish and other aquatic organisms. These incidents are known as Harmful Algal Blooms (HABs). Cyanobacteria produce three broad classes of toxins known as hepatotoxins, neurotoxins, and dermatotoxins that affect the liver, nerves, and skin, respectively (Sivonen and Jones 1999; Williams et al. 2007; Burns Jr 2008). In addition to toxic effects, general irritation can occur upon contact (Chorus and Bartram 1999). Swimmers and anglers have complained of rashes after coming into contact with blooms (Steidinger et al. 1973). The World Health Organization (WHO) has set a drinking water “provisional consumption” limit of 1 mg/L for one type of cyanotoxin, microcystin-LR, a toxin produced by several types of cyanobacteria, including Microcystis species (Chorus and Bartram 1999), and the U.S. EPA has issued informal health advisory guidelines for 1.6 mg/L microcystin-LR in drinking water for school aged children and adults (EPA 2015a).
For recreational waters, the U.S. has no guidelines, but the EPA has recently drafted recommendations for microcystins at 4 ug/L (see section 2.4.5). The WHO considers 10-20 µg/L microcystin-LR to have a moderate probability of adverse health effects for a 132-pound adult that ingests 3.4 oz. contaminated water. A 30-pound child would need to ingest less than 1 ounce for the same risk (Chorus and Bartram 1999). As cyanobacteria concentrations increase, so does the potential for people to ingest toxins at levels that can cause adverse effects. Scums produced by some species such as Microcystis and Anabaena are particularly hazardous. They contain high levels of toxin, so it is important for the public and their pets to avoid exposure to them (Chorus and Bartram 1999). Four summary references on HAB by Steidinger et al. 1999, Burns Jr 2008, Williams et al. 2007, and Abbott et al. 2009 are recommended reading on this subject.
The St. Johns River and its tributaries are impacted by excess nutrients in runoff and wastewater (section 2.3). Nutrients, including nitrogen- and phosphorus-based chemicals contained in garden, lawn, and agricultural fertilizer, are common causes of impaired waters in the LSJR and are crucial contributors to freshwater algal blooms.
High levels of nutrients lead to phytoplankton growth and eutrophication, causing the ecosystem to become unbalanced with increased loading of organic matter to the system as a result (NRC 2000). Thus, when nutrient levels are high and other appropriate conditions exist, the possibility of harmful algal blooms increases. Cyanobacteria growth rates and species distributions in an ecosystem are highly dependent upon light, temperature, and salinity. As a consequence, proximity to the mouth of the river (due to salinity levels), temperature fluctuations, color of the water, and the presence of other phytoplankton all determine whether an algal bloom will occur and which species will predominate. Rainfall also influences HABs; periods of low flow during drought increase the likelihood of algal blooms in the freshwater reach (Phlips et al. 2007), while high flow and hurricane rain events increase the likelihood of less concentrated but more widespread blooms in the downstream, Jacksonville reach of the river (Hendrickson 2013).
Nutrients promoting algal blooms also come from leaking septic systems, livestock, industry and runoff during and after heavy rain events. However, interesting work by Piehler et al. 2009 indicates some types of cyanobacteria can themselves increase nitrogen in waterbodies. During nitrogen fixation, a biological process, atmospheric nitrogen is taken up and used for growth by some species. The nitrogen is ultimately released into the water in forms that are more usable by biota that cannot use atmospheric nitrogen.
The question often arises about whether harmful algal blooms occurred historically and whether current blooms are a natural occurrence. Burns has this to say (Burns Jr 2008):
“Although there is little doubt that the phenomenon of cyanobacterial blooms predates human development in Florida, the recent acceleration in population growth and associated changes to surrounding landscapes has contributed to the increased frequency, duration, and intensity of cyanobacterial blooms and precipitated public concern over their possible harmful effects to aquatic ecosystems and human health. Toxic cyanobacterial blooms in Florida waters represent a major threat to water quality, ecosystem stability, surface drinking water supplies, and public health.”
Interestingly, algal blooms may have increased after successful eradication efforts to control the highly invasive water hyacinth in the 1970s and 1980s. In the past, hyacinth shaded much of the water column and limited algal growth. Reduction in the water hyacinth may have contributed to the change from a floating aquatic plant system to an algal-dominated system in the LSJR (Moody 1970; Hendrickson 2006; Hendrickson 2008).
2.4.2. Cyanobacteria in Florida and the LSJR
Anabaena circinalis and Microcystis aeruginosa (Figure 2.17) are two of the most widely distributed freshwater cyanobacteria species in Florida that generate HABs (Steidinger et al. 1999; Williams et al. 2007; Abbott et al. 2009). Some of the other potentially toxic cyanobacteria that are known to bloom in Florida waters include Cylindrospermopsis raciborskii (reported as a possibly recent invasive species (Chapman and Schelske 1997)), Anabaena flos-aquae, Aphanizomenon flos-aquae, and Lyngbya wollei (Steidinger et al. 1999; Burns Jr 2008; Abbott et al. 2009). Extensive statewide sampling by Florida biologists in 1999-2000 showed that 88 out of 167 samples, representing 75 individual waterbodies, were found to contain potentially toxic cyanobacteria (Williams et al. 2001; Burns Jr 2008). Most bloom-forming cyanobacteria genera were distributed throughout the state, but waterbodies, such as Lake Okeechobee, the LSJR, the Caloosahatchee River, Lake George, Crescent Lake, Doctors Lake, and the St. Lucie River (among others) were waterbodies that supported extensive cyanobacterial biomass. Seven genera of cyanobacteria were identified in the statewide samples, with Microcystis (43.1%), Cylindrospermopsis (39.5%), and Anabaena (28.7%) the most frequently observed, and in greatest concentrations (Williams et al. 2001; Burns Jr 2008). In the same 1999-2000 survey, 55% of the samples in the LSJR basin contained the genus Anabaena, 53.9% contained Cylindrospermopsis raciborskii, and 47.6% contained the genus Microcystis (Williams et al. 2001; Burns Jr 2008), though it should be noted that many other species reside in the LSJR.
In 2005, major blooms in the LSJR affected areas north of Crescent City to Jacksonville and caused large spikes in cyanotoxins and fish die-offs. The primary species was Microcystis aeruginosa (Williams et al. 2006; Williams et al. 2007). In an unusual series of events in the LSJR from mid‐May through June of 2010, cyanobacteria blooms grew in great abundance in the freshwater reaches of the LSJR, beginning with blooms of Aphanizomenon cf. flos‐aquae, which until then had never been recorded as the dominant species in the LSJR. With an increase in river salinity due to extended periods of reverse flow, the Aphanizomenon bloom decayed and was replaced by Microcystis, Cylindrospermopsis, Anabaena, and Pseudoanabaena (FWC 2010). Analyses for cyanotoxins, which are toxic chemicals produced by cyanobacteria, indicated large spikes of a microcystin in the river water in late May and June and elevated levels of Cylindrospermopsis in mid‐July through September (Hendrickson 2011).
Other cyanobacteria identified in the LSJR in 2012 by the SJRWMD field observation team include Anabaena spiroides Oscillatoria limosa, in the Ocklawaha River in 2012, as well as Planktolyngbya limnetica and Planktolyngbya tallingi in Crescent Lake (LSJR TAC 2012).
Identification and quantitation of cyanobacteria and their toxins in the LSJR can be difficult, expensive, and time-consuming, though in recent years, there has been an expansion of different methods and approaches (Williams et al. 2007; Burns Jr 2008). The most consistent and complete data that reflect phytoplankton growth over many years are measurements of chlorophyll-a. Chlorophyll-a is a light-harvesting pigment used by photosynthesizing organisms. Elevated phytoplankton concentrations, including cyanobacteria, are accompanied by elevated chlorophyll-a concentrations so chlorophyll-a is often used as an indicator for HABs.
2.4.3. Chlorophyll-a Thresholds and Data Analysis
Chlorophyll-a values are used to determine relative phytoplankton abundance. Each water body is unique with respect to flow, shape, and water chemistry, all which affect phytoplankton growth and therefore also chlorophyll-a levels (DEP 2013d). Because salinity is a critical factor in cyanobacteria growth, it is useful to examine chlorophyll-a in different river regions. The marine/estuarine reach discussed in this report extends from the mouth at WBID 2213A to WBID 2213H, and the freshwater region extends from WBID 2213I upstream to WBID 2213N at the confluence of the Ocklawaha River (Figure 2.1).
Criteria and threshold values
Streams with chlorophyll-a concentrations that are below 3.2 µg/L are biologically healthy; however, some types of streams are stable and healthy at higher levels of chlorophyll-a/L. Therefore, a number of DEP chlorophyll-a impairment thresholds exist for Florida waterways ranging from general criteria to site-specific criteria. For example, the impairment threshold for estuaries and open coastal waters is 11 µg chlorophyll-a/L (annual geometric mean) (DEP 2013d). However, the marine/estuarine reach of the LSJR has an even lower, more stringent, site-specific chlorophyll-a criterion of 5.4 µg/L for long-term (7-year) annual averages (DEP 2013h; DEP 2016a). Thus, 5.4 µg/L is the threshold criterion we utilize in this report for the marine/estuarine reach (which is most appropriately compared to 7-year annual averages).
For freshwater, the general impairment threshold in Florida is 20 µg chlorophyll-a/L (not to be exceeded more than once in a three year period), based on annual geometric means (DEP 2013d; DEP 2016g). However, the DEP uses a criterion of 40 µg chlorophyll-a/L not to be exceeded more than 10% of the time (or for more than 40 days) for the chlorophyll-a threshold for the Lower St. Johns River Basin (Magley and Joyner 2008; DEP 2014a). For this River Report, both freshwater criteria are used – the general 20 µg/L annual geometric mean threshold is used for general assessment of the freshwater regions of the river, and 40 µg/L chlorophyll-a/L is used as a threshold of bloom status when discussing individual water segments (WBIDs).
In this River Report, the current status and time trends of chlorophyll-a are examined in different ways. Both the marine/estuarine and the freshwater regions are presented as annual chlorophyll-a averages for direct comparison between the two regions (Figure 2.18). The marine/estuarine data are then displayed as annual averages as well as 7-year annual averages and are compared to the 5.4 µg/L chlorophyll-a criteria (Figure 2.19A). The freshwater data are presented as annual geometric means for comparisons to the 20 µg/L chlorophyll-a threshold (Figure 2.19B). To show the spread of the data, including the high and low values, box-and-whisker plots are presented for both the marine/estuarine and freshwater regions (Figure 2.20). Trends over time of annual average concentrations were investigated by using the Spearman Rank 1-tailed test at p < 0.05.
Data acquisition and processing
All data were obtained from the DEP STORET. STORET is the statewide environmental data system containing water quality, biological, and physical data. Method 10200-H was used to analyze chlorophyll-a that was corrected for pheophytin, which is a form of degraded chlorophyll. Only stations in the mainstem or near the mainstem in major tributaries, such as the Ortega River and Julington Creek, were included. Data were reviewed for quality and data points were discarded when samples appeared analytically compromised (contaminated blanks, poor recovery, poor replication, etc.) or were missing important information. All samples with qualifier codes K, L, O, V, Y, or ?, which indicate different data quality issues, were eliminated. If a reported value was below the method detection limit (MDL), it was used even if flagged. One-half the MDL was used for samples reported as “nondetect.” In a small number of cases, the MDL was estimated by determining the MDL reported most frequently for other samples during the same year. When routine and integrated vertical samples were obtained at the same time, the integrated samples were used for analysis in this report.
2.4.4. Current Status and Trends
Based on annual averages, the freshwater regions have shown consistently higher concentrations of chlorophyll-a compared to the marine/estuarine regions (approximately two to seven times higher per year; Figure 2.19), which is not unexpected since the phytoplankton that cause algal blooms in the St. Johns River are freshwater species. Using aggregated WBID data, the decrease in chlorophyll-a concentrations when moving from the most upstream freshwater WBID N to the end of the St. Johns River at the Atlantic Ocean (WBID A) is evident (Fig. 2.19B).
To assess the chlorophyll-a levels in the marine reach, the annual averages are compared to the impairment criterion value of 5.4 µg/L (Figure 2.20B). The yearly data (blue diamonds) show statistically significant decreases in chlorophyll-a concentrations over the past six years, even reaching below the 5.4 µg/L threshold the past three years. However, the 5.4 µg/L threshold is meant to be compared to long-term, 7-year averages, not individual years. The 7-year averages (black squares) have consistently been above the 5.4 µg/L threshold, thus the marine reach does not meet the chlorophyll-a criterion. Therefore, while the data have shown decreasing chlorophyll concentrations in the marine reach, which is promising, the 7-year long-term averages are all above the 5.4 µg/L target (Figure 2.20B). However, the lack of recent data for WBIDs C and F constrain the overall interpretation.
The freshwater annual geometric means for the five most recent years are below the 20 µg/L threshold that this River Report uses for comparison (Figure 2.20A), and these data also show a statistically significant downward trend in chlorophyll-a concentrations over the past 6 years, which looks promising regarding reduction in freshwater blooms. While Figures 2.19 and 2.20 are useful for trends and comparison to threshold values, they do a relatively poor job of representing the data sets per year. Furthermore, geometric means tend to be lower than arithmetic means (compare Figure 2.19A to 2.20A). Therefore, box and whisker plots are presented to show the spread of the data in 25% increments as well as the median value, the highest values, and the lowest values (Figure 2.21; see circled inset). The highest freshwater values for the past five years and the highest marine values for the past four years are the lowest high-values for the 20 years analyzed for each segment of the river (Figure 2.21 A and B, respectively).
While the trend of lower chlorophyll-a concentrations is encouraging, the aggregation of the data for the large freshwater region has its limitations. Not only is the freshwater region large, but it is also complex, and there are still sections of this region (WBIDs; Figure 2.1) that are experiencing elevated chlorophyll-a levels, and therefore it is worth discussing WBID-specific information instead of strictly the aggregated data from the freshwater portion. First, figure 2.19B illustrates historic differences between the WBIDs, including among the freshwater WBIDs, with K-N noticeably higher than H-J. Second, the points below illustrate more in-depth localized analysis of data for freshwater WBIDs. These specific examples are intended to highlight issues with data limitations as well as highlight how specific regions of the freshwater portion endure particularly elevated chlorophyll-a levels, including levels which indicate blooms.For 2012 and 2013, 5 of the 6 annual chlorophyll-a geometric means of freshwater WBIDs K, M, and N were above the 20 µg/L threshold.
- Over the most recent 4 years (2013-2016), WBIDs K, M, and N combined have an almost 2- fold higher chlorophyll-a average compared to combined WBIDs H, I, J, and L.
- For 2016, 8 of 22 samples from WBID N exceeded the site-specific chlorophyll-a criterion of 40 µg/L. These 8 samples spanned May to September, and include all 5 samples from July 13 to September 7, indicating a prolonged bloom. Furthermore, the annual geometric means for this WBID ranged from 20.4-22.7 µg/L for 2012, 2013, and 2016, slightly over the 20 µg/L threshold for each of those years. Therefore, WBID N exceeds the two different thresholds used in this River Report, including the state criteria.
- The DEP has recently analyzed chlorophyll-a data from WBID K (Racy Point), which is a location that it considers to be a “worst-case WBID”, along with nearby Dancy Point (WBID L), and reported the number of days per year that Racy Point experienced a nuisance bloom (designated as chlorophyll-a values >40 µg/L). That analysis reports a trend of decreasing days per year (from 1995-2013) when Racy Point is >40 µg/L chlorophyll-a (DEP 2014a). While that analysis shows a decrease in the longevity of these blooms at this location, it still has recurring blooms. In that analysis, chlorophyll-a values derived from continuous measurements at Dancy Point (Using USGS data, which are not deposited in the DEP STORET database) are compared to the values of chlorophyll-a measured from discrete samples at Racy Point (that are in the DEP STORET database). That analysis shows many days during spring 2013 when Dancy Point (WBID L) was above the 40 µg/L threshold according to the continuous measurements, but these types of data are not in the DEP STORET database. This lack of WIBD L data for 2013 is an example of limitations of figures 2.19-2.21, and illustrates the problem of our analysis missing key algal bloom events in the river.
- WBID K has only three data points in the database for 2016, with a high value of 43 µg/L and geometric average of 22 µg/L.
Chlorophyll-a data relate to abundance of all of the phytoplankton present. When high concentrations of chlorophyll-a are specifically from toxic cyanobacteria, then concerns about water toxicity are elevated. In October 2013, for example, the St. Johns Riverkeeper reported that microcystin (a cyanotoxin) concentrations in two LSJR-associated samples were >2,000 µg/L. These concentrations are more than 500 times the new USEPA draft recommendations for recreational waters (see New EPA recommendations section below; EPA 2016a; EPA 2017), and are at a level that the World Health Organization classifies as posing a very high probability of acute health effects from recreational exposure (Inclan 2013; Patterson 2013; St. Johns Riverkeeper 2013a; EPA 2016b). One of the toxin-sampling sites was in Doctors Lake (also see Figure 2.18 for photos of blooms in Doctors Lake), which feeds into the St. Johns River and the other was in the more estuarine region of the river (at Jacksonville University). While both of these samples had very high levels of toxins, the chlorophyll-a samples in the STORET database for Doctors Lake that same month were less than 40 µg/L, and none of the 2013 estuarine samples in the analyzed data set were >40 µg/L chlorophyll-a. Thus, there were blooms that are not evident from looking at the chlorophyll-a levels from the STORET database, which illustrates another limitation of the chlorophyll-a-based analysis in this report.
The above discussion demonstrates that the datasets we are using have limitations in that there are recorded instances of high chlorophyll-a levels that are not captured in our analysis, and there are toxic algal bloom events that have occurred that are not represented in the chlorophyll dataset. Furthermore, the river sampling locations for chlorophyll-a are largely in the middle of the river channel, and it is known that Microcystis blooms can be concentrated along the shore and in coves due to the ability of these cyanobacteria to float and be pushed by the wind. Thus, some wind-driven blooms or elevated chlorophyll-a levels are missed with routine sampling at fixed locations. Sampling protocols can also miss more dense algae at the surface of the water. Finally, the geographical region analyzed also affects the interpretation. While the data here were grouped as either freshwater or marine/estuarine for the main analysis (Figures 2.19-2.21), there are individual locations in the river and its tributaries that are particularly problematic regarding algal blooms.
2.4.5. Summary and Future Outlook
The past few years have shown mean and median chlorophyll values lower than the threshold values, which is promising. However, the marine reach is still not in compliance, and for both freshwater and marine segments of the LSJR, the number of chlorophyll-a exceedances and the annual appearances of blooms (including toxic events) indicate significant impact from phytoplankton, including cyanobacteria. Therefore, the STATUS of the LSJR with respect to algal blooms is considered unsatisfactory, and the TREND is unchanged.
Much of the outlook for algal blooms is closely tied to that for nutrients. Reduced nutrient loading may be lowering concentrations of some forms of nutrients in the mainstem (Section 2.3), and perhaps an accompanying reduction in algal blooms is starting to be observed. However, more years of data will be needed to understand whether this is a stable trend or a short-term phenomenon. Furthermore, the number of samples obtained from the database over the past few years is reduced compared to previous years, which makes trend analysis more complicated, particularly if algal blooms are missed due to decreased sampling. The next few years will be critical in determining whether the considerable effort and expenditures to reduce nutrients in the LSJR are sufficient to limit algal blooms.
Complexity of the system
The freshwater and marine/estuarine sections analyzed above are large geographical areas and thus give a big picture view of the LSJR over time. It is important to consider the complexity of the LSJR ecology and the difficulty in establishing healthy benchmarks and natural trends in a system with physical, chemical, and biological characteristics that naturally vary so widely in space and time.
New EPA recommendations
In December 2016, the USEPA published draft recommendations for national water quality criteria for microcystins and cylindrospermopsin with the goal of protecting swimmers and others engaged in recreational activities in natural waters (EPA 2016a; EPA 2017). The swimming advisory levels, which are not to be exceeded on any day, are 4 ug/L for microcystins, and 8 ug/L for cylindrospermopsins. For assessment purposes, both toxins also have a waterbody impairment recommendation of not to exceed the above values ‘more than 10% of days per recreational season up to one calendar year’. The comment period for the draft recommendations ended March 2017 with no published timeline for finalizing the recommendations.
2.4.6. Recommendations for Research
As discussed above, some bloom events are not represented in the FDEP STORET database, so our understanding of frequency, duration, and locations of blooms is not comprehensive. Additional monitoring of chlorophyll-a at locations known to have recurring blooms, combined with associated nutrient, temperature, and toxin data, could help form a clearer picture of these blooms and the extent to which they exist, how toxic they become, and determine drivers of bloom formation, such as nutrients and temperature. Laboratory, mesocosm, and in situ studies that analyze growth rates, toxin production, and bloom collapse of HAB cyanobacteria isolated from the LSJR as a function of varied nutrients, salinity, and temperature are essential to understanding blooms of the LSJR.