Recent cyanobacteria and algae blooms in Island Park reservoir, and an introduction to the detected shift in the timing of downstream seasonal water-temperature increases

Brief background: There are three genera and many species of cyanobacteria; only a handful of these produce a toxin harmful to mammals. Blue-green algae (or BGA) is the common name for cyanobacteria, which is a single-celled bacteria that photosynthesizes and is neither an algae nor a plant. All plants, algae, and cyanobacteria use chlorophyll A, a pigment used to capture light for photosynthesis. In addition to chlorophyll A, freshwater cyanobacteria also utilize phycocyanin for photosynthesis, which is a bluish pigment that gives them their color. When cyanobacteria begin to die and disintegrate, this pigment may color the water a distinctive bluish hue. When algae and plants begin to die and disintegrate they can give the water a dark color. Individual cyanobacteria and algae can have a short life span, but a bloom, which is a colony or aggregation of individuals, can last for days or weeks depending on environmental conditions, like nutrients, temperature, and flow. Using our data, the most-likely explanation for the greyish color of the water last Monday (the 11th) is the combination of dying and decaying cyanobacteria and algae in the reservoir, which fall to the bottom and get pulled out through the outlets near the bottom and delivered downstream.

What we see in the data: The sondes detect changes in content of cyanobacteria, algae, and plant biomass by testing for the presence of their photosynthesizing pigments chlorophyll A and phycocyanin. We saw slight daily increases in cyanobacteria on Thursday and Sunday last week, then a small increase on Monday the 11th, and a smaller, slight increase on Tuesday the 12th (Figures 1 and 2 below, bottom panel in each). Cyanobacteria have a short life span, approx. 12 hours, so if we're seeing an increase in living cyanobacteria as early as Thursday, there is dead and decaying cyanobacteria quickly accruing at the bottom of the dam and immediately below the dam. Recall decaying cyanobacteria turn the water a bluish hue. It's possible that cyanobacteria is not only producing inside of the reservoir, but may also be producing in eddies or along shallow slack-water banks below the dam.

Is this of ecological concern for the fish? The toxin that some cyanobacteria produce is not harmful to fish, so that's not the issue. However, a significant cyanobacteria bloom can be harmful to fish if it decreases DO (dissolved oxygen) below their metabolic requirements during warm weather (warmer weather speeds their metabolism, which makes them require even more oxygen). These blooms occur during warm weather, when oxygen levels are already seasonally low, but the kicker is when the blooms suddenly die, their decomposition process can use up a lot of oxygen, which could drive DO levels down to stressful low levels for fish. This may cause a fish kill if the fish cannot find refuge in water with higher DO. There are layers of higher DO in the reservoir near the surface, so this isn’t currently a concern. Also, thanks to Fall River Rural Electric running their industrial aerators this year, DO has been above 7 mg/L all week in the river immediately below the reservoir (7 mg/L is shown by the dashed line in the DO panel in Figures 2 and 3), which is comfortable for trout this time of year.

The other way a bloom can impact DO for fish is if the bloom, specifically cyanobacteria, is so extreme that the individuals form massive colonies that look like a thick layer of paint on the water, which can block out light, preventing the usual plants from photosynthesizing, which may drive DO too low. We haven't seen evidence that a bloom has driven DO below the Rainbow trout habitat requirement in the sonde data this year, nor in the previous two years. What’s more, any potential harm to fish will most likely occur in the reservoir where algae and cyanobacteria have the opportunity to accumulate, not in the river. The bloom (a large and prevalent aggregation of individuals) isn't in the river and can't grow in the river, at least as long as flows are high. The effects in the river are limited to coloration from the dead cells and a few dispersed individuals.

Is this a health risk? The toxin produced by some blooms is harmful to birds and mammals if ingested or during primary contact recreation, like swimming or water skiing, where one may ingest water accidentally or one is submerged long enough that a skin reaction can occur. Ingestion can cause vomiting, nausea, and a host of other symptoms, but no human deaths yet have been associated with algae or cyanobacteria blooms. Numerous dog deaths across the U.S. have, though. Of those cyanobacteria species that can be toxic, most of the time they aren't... they often don't produce the toxin and it isn't understood why they do, when they do. Also, there is very little evidence pointing to other signs that can be used to infer whether the toxin is present (aside from someone's dog dying after going for a swim). The only way to know for sure is to wait for a lab result. HFF contacted Troy Saffle, the Water Quality Manager for the Idaho Falls Region DEQ, on Wednesday July 13 and they will now incorporate Island Park reservoir into their HAB (harmful algal bloom) sampling this season. It is common knowledge that visual evidence of a large aggregation of cyanobacteria, which may look like a paint slick, provides reason to take warning. Even though we haven’t seen any aggregations thick enough to look like an opaque paint slick, toxic cyanobacteria have been confirmed in IP reservoir in the past, so it's not worth risking it: fishing should be fine, but until we report the results from DEQ’s lab tests you should rethink going for a long face-submerged swim, or letting your dog lap up the water.

Summary and mention of shifting seasonal water temps: We've seen evidence of small cyanobacteria blooms in the reservoir in the sonde record since 2014, but last week was comparatively early in the year to see one. Cyanobacteria blooms thrive if there are enough nutrients to feed on, if temps are warm, and if flow is relatively low so that their colonies aren't broken up. The reservoir acts more like a lake than a river, even when outflow is as high as it is right now, and the temps have been very warm for this time of year and it's been unrelentingly sunny. Outflow from the reservoir has been above 65˚F since Wednesday July 6th compared to this week in 2014, when we hadn't yet broken 58˚F and wouldn't reach 65˚F until a 10-day period from July 25th to August 5th. Therefore, blooms this early may be more prevalent in the years to come if we continue to have earlier warm and sunny weather in combination with dry conditions. Climate is a major factor in these blooms and keeping the reservoir as full as possible helps quality of downstream water by preserving thermal stratification for longer into the warm season. Thermal stratification in the reservoir means cooler, denser water has sunk to the bottom and the outlets near the bottom draw this cooler water downstream. What we can expect for the remainder of the season is water temperatures will probably cool a bit once irrigation delivery slows and stratification sets back up in the reservoir. As mentioned in the Gill Lice Study information sheet, much cooler water temperatures will also slow the reproduction of gill lice. I will prepare a follow-up blog, coming soon, that will detail the intriguing temperature shifts through time we’ve detected from our data and how these temperature shifts may be driving changes in the river.

Figure 1 Depth profiles from Island Park Reservoir, July 8th 2015

Figure 2 Select sonde data from the IP East sonde downstream of the power plant outflow on the eastern bank from last week, July 6 through July 12.


Figure 3 Select sonde data from the IP West sonde downstream of the USBR gates outflow on the western bank from last week, July 6 through July 12.


The paragraphs above summarize our major conclusions from the last week’s data. The paragraphs below briefly step through each of the three data sets we used to draw those conclusions for those interested parties.


First, let's look at the depth profiles from Friday the 8th, Figure 1. Recall the depth profiles represent a single snapshot of physical and chemical parameters in the reservoir's water column. You see the profiles from in front of the gates intake on the top row and from in front of the power plant intake on the bottom row. In each panel, the top is the reservoir surface, and the bottom of the graph is the reservoir bottom in that particular location. You see living cyanobacteria in the top 5 meters in both locations and slight evidence of algae near the surface, too. There are large spikes when the sonde hits the bottom, which is layer of accumulated deposition, and the spikes indicate recently dead individuals whose photosynthetic material haven’t yet broken apart during decomposition and still registers on the sonde. These would represent only a portion of the total material on the bottom from cyanobacteria and algae, since material that has undergone more advanced decay wouldn't register on the cyanobacteria and chlorophyll A probe, but would register on the turbidity probe.  

In the turbidity record, you can see how conspicuous the living cyanobacteria and algae are by those increases in the top 5 meters at both locations. The recently dead individuals will show up just as well as the living in optical turbidity data, so, using the cyanobacteria and chlorophyll data at the bottom of the reservoir, you can infer that a large portion of the observed turbidity at the bottom is due to dead and decaying cyanobacteria and plant material. The intakes for both discharge outlets are near the bottom of the reservoir, so they are pulling a mixture of decayed plant, algae, and cyanobacteria material as well as actual mineral sediment. We also know this to be the case from the water samples we tested for Total Organic Carbon last year, where we found that approximately half of the suspended material immediately downstream of the dam during this time of year is mineral sediment and half is decaying organic material, like dead plants and output from chemical reactions, all of which will be taken up and used by the downstream food web.

Now let's look at the sonde records from IP East (downstream of the power plant outflow, Figure 2) and IP West (the gates outflow, Figure 3) from mid-day Wednesday July 6th to mid-day Tuesday July 12th. You see water quality parameters through time: temperature, dissolved oxygen (DO), turbidity, and in the bottom panel, cyanobacteria content denoted by the bluish line and chlorophyll A by the green line.
During the represented period there was approximately 1600 cfs in total being delivered downstream with the power plant running at full capacity, releasing 960 cfs of the total discharge, so we can assume that downstream water quality is an approximately-even mixture of what's in these 2 records. Notice, temperatures were warm, between 18-19 ˚C (64-66 ˚F) all week, and recall the discussion above that these are unseasonably warm water temperatures, which is a key driver of algae and cyanobacteria blooms. Notice DO is consistently above 7 mg/L (dashed line in the DO panel, Figures 2 and 3) since FRRE has been running their aerators. Skip the turbidity record for right now. Notice the daily increases in chlorophyll A and cyanobacteria and the corresponding timing of the main increase with the time an odd color was noticed in the river. Then look back at the turbidity record and notice that there has been a steady and almost linear increase all week, which corresponds to what we would expect to see as the reservoir is drawn down and deposited material on the bottom becomes mobilized, and then a slight increase in the slope of that line beginning on Monday the 11th, due to the algae and cyanobacteria bloom. What we didn't see in the turbidity record was a large and sudden increase on Monday the 11th, only a slight increase of less than a couple FNU. The point is a large change in the color of the water may not correspond to an equally large change in turbidity from mineral sediment, but may correspond to a significant perceived change in clarity due to the color change itself. Think of a couple drops of red food coloring and how drastically that will change the color in a cup of water while very minimally impacting the actual clarity of the water. Without a change in the color of the water to a bluish hue from decaying cyanobacteria hardly anyone would notice a < 2 FNU change in turbidity, it's barely perceptible. The sonde record shows that there was a minimal change in turbidity, but does show a strong change in living or very recently dead cyanobacteria, and as discussed above, if we infer the amount of dead and decaying cyanobacteria that is being sucked out from the bottom of the reservoir along with that recently dead cyanobacteria, that would provide a much greater change in color (to a bluish hue) than increase in turbidity. And this is what we see in the turbidity data: a modest increase in turbidity on Monday the 11th that corresponds to the increase in cyanobacteria and algae being produced in the reservoir and probably being produced in conducive areas downstream. It deserves to be said, that while a drastic change in color may not correspond to an equally large change in actual mineral sediment mobilization, an increase in turbidity along with a color change may interfere with fishing since conditions have changed for the fish, even slightly, which inspires a change in behavior.