Case of the Missing Lake Merced Hematite


It was a bright and sunny day out on the Lake and we had just nailed our big clue:


These results indicated… Oops…! I’ve gotten ahead of my story. Let me start at the beginning.

It was a bright and sunny Sunday, March 11th, out on Lake Merced. We were making a second survey of the lake to find out more about the rust red material we thought might indicate Iron Bacteria. For this part, you may want to take time to look back at my last Lake Merced Post. You should also look at the aerial photographs Cris Benton took during this hike.

Our first objective was to document the area around Sample Site 01, where we had found the first showing of rust red material in the shallow water along the edge of the lake. Our second objective was to see if similar shallow areas of the lake shore showed the rust red material, and, if so, to map those areas.

Susan Schwartzenberg documented the sampling of the rust red shallows near the Trail Bridge:

Susan - Red Surface Sample 01

We were careful to take the sample only from the surface and only of the rust red material; or, only from the bottom mud of the shallows. Here are the three samples we took from Sample Site 01 – Trail Bridge area:


For the moment, I am going to focus on the surface samples of the rust red material. Shown below is my set-up for testing whether the rust red material is Hematite or not;

In this test, hydrochloric acid [HCl] is mixed with the sample material. If the sample material contains Iron{+++], the HCl displaces any other molecule bound to the iron with the chlorine [Cl]. The resulting compound is soluble in water and is yellow.

This is the reaction: [It is a simple chemical test that anthropologists and archeologists use for determining the metal in corroded artifacts.] Hydrochloric acid reacts with Hematite, displacing the oxygen and producing FeCl3 which is brilliant yellow in solution.

Fe2O3 + 6HCl ? 2FeCl3 + 3H2O [aquous solution = YELLOW]

In the video shown below, the test is being performed with the rust red surface material from the sample. Notice how quickly the rust red material dissolves in the solution of HCl and the solution turns from clear to greenish-yellow.

Now, if you look back to the first photograph, the two pictures of the tube in which the test was carried out, you can see the clear HCl before the sample was added and the clear yellow solution after the sample was added.

This confirms that the rust red material sampled was Hematite.

In my original post, I postulated that the red rust material was Hematite, produced from Ferrous Sulfate by Iron Bacteria. This confirms the Hematite part of the hypothesis. We also performed the test on red rust material collected by floating a small sheet of paper just below the surface [see the photograph of samples].

We re-examined the bottom sample for various bacteria and found many individual cells, double cells, and chains of bacteria like the following:


These bacteria are about 5-microns long, about the size and shape of Thiobacillus ferrooxidans, a common species of iron bacteria. We found them in the bottom mud of the shallows near the Trail Bridge of Sample Site 01 and we found them associated with the red rust material on or near the surface of the shallows.

Unfortunately, taxonomic identification of bacteria by photomicrograph is not very reliable. We have not established the genus and species of these bacteria, we have not established that they actually produced the Hematite, nor have we established that they produced the Hematite by oxidizing Ferrous Sulfate from fertilizers or other sources around or near the lake.

Here, we see that the bacteria are slightly motile. The wiggling oval object is a Dinoflagellate. The slowly moving chain is composed of the bacteria. Chains like this are formed when division is rapid.

In the sample from which this videomicrograph was taken, there were many rod-shaped bacteria like this, in singles, in pairs, and in chains of various lengths.

I set aside an individual sample from the Sample 01 bottom sample as a control, then added 3.1-grams of Ferrous Sulfate to the jar. Then I set the jar in the window to incubate for awhile. The jar of bottom sample contained very few flakes of the rust red material that we have identified as Hematite.

Any Hematite that appears in the jar upon incubation for a few days is likely to have been produced from the Ferrous Sulfate added, possibly by the rod-shaped bacteria. This will be confirmed if the individual control sample set aside, to which no Ferrous Sulfate was added, shows no Hematite.

This experiment still does not confirm that the rod-shaped bacteria are either Thiobacillus ferrooxidans or some other Iron Bacteria. This can be confirmed by culturing the bacteria and performing various diagnostic tests, or, more simply, by DNA identification. The region of DNA that codes for 16S RNA is sometimes referred to as a “DNA Barcode”, since it is commonly used for identification. We are presently looking for ways to perform such identification, since we are not equipped for DNA amplification and sequencing.

The other objective of our second visit to Lake Merced was to map the Hematite in the lake shore shallows. If the Hematite was produced from Ferrous Sulfate in fertilizers used on the Golf Course, then we would expect to see the Hematite in all the shallows near or across from the Golf Course run-off areas.

We went to a site just below the Golf Course [now designated Sample Site 03 N37.72641 W122.49857], directly across the narrow neck of Lake Merced from Sample Site 01. This was a similar Fishermen’s Bridge with shallows between it and the lake shore. The bank down to the shallows was steep and directly below the Golf Course green.

Examination showed no sign of Hematite in the shallows. This shallow area has a dense growth of Tule Reeds, but no Duckweed and no red rust material. It was more similar to Sample Site 02, a few hundred yards to the west.

This was a startling and unexpected finding. Fertilizer-rich run-off from the Golf Course would have to reach Sample Site 03 before it could reach Sample Site 01. At this point our hypothesis looked shaky.

Hiking back from Sample Site 03 to the parking lot, we walked along one of the east-west Golf Cart Paths that passed directly to the north of the Club House. Suddenly we saw rust red mud along the edge of the Golf Cart Path. The area was wet, as if it had been recently watered.


I was out of sample vials, but used a pipette to collect about 2-ml of this rust-red mud and a little of the water with it. When we got to the car, I transferred it to a spare vial borrowed from Cris Benton.

When I tested the Sample Site 01 TrailBridge bottom sample for Hematite, I also tested the sample from the Golf Cart Path. It was a smaller sample, so the reaction was not as strong, but it was clearly positive for Hematite.

What to make of this? Hematite in Sample Site 01 across the narrow neck of the lake from the Golf Course. Hematite in a sample from the Golf Cart Path near the Club House. No Hematite in either of two samples on the Golf Course shoreline, Sample Site 03 directly below the Golf Course and Sample Site 02 to the west of the Golf Course.

A possible hypothesis is a common watering system. I understand that the water used in the park around the lake is reclaimed water. Such water is not purified, so fertilizer components might be in the water, particularly Ferrous Sulfate.

Consistent with this hypothesis was our finding of a small shallow pond somewhat removed from the lake, but near Sample Site 01.


Notice that the small pond is rich with the red rust material. Subsequent tests confirmed Hematite.

This pond is well away from the lake shoreline and somewhat higher. Since the lake has been going down, not rising, it is not clear how this pond could have resulted from Lake Merced water.

There are, however, watering connections in the area. It may well be that water in a common watering system has been applied near the Trail Bridge Site 01, the Club House Cart Path, and this pond. It is unlikely that the water itself is rich in Hematite, but it is more likely that the reclaimed water is rich in fertilizer components such as Ferrous Sulfate. This would encourage the growth of Iron Bacteria, producing the Hematite.

Science, unlike TV detective stories [but probably more like real-life CSI], is filled will uncertainties, false leads, wrong hypotheses, chance observations, and – sometimes – right conclusions in the end. Best to be driven by the data, not by the favored hypothesis.

We still have a lot to do. More mapping of Hematite on the shoreline and away from the shores. Tracing the watering system. Possibly checking the watering system. Taxonomic identification of the bacteria.

In the end, we will also have to face the practical question of what this means about the health and well-being of Lake Merced. So far, the bottom of all sample sites has been largely barren. Except for the rod bacteria, there are few microorganisms of any kind and certainly not the rich mat communities I have seen in the shallow areas of other lakes and streams. There may be no connection, however, to the presence of Hematite or Iron Bacteria or fertilizer run-off.

It is still a mystery.

ADDENDUM: Lake Merced Sample Site 01 TrailBridge bottom sample had no floating Hematite [rust red material]. Testing the water from above the mud in the jar for Hematite showed little reaction. This sample was carefully taken from the mud and water just above the mud in the pond at TrailBridge. It showed a fair population of the rod bacteria similar to Thiobacillus ferrooxidans [Iron Bacteria].

Last week I set part of the sample aside as a control, then I added 3.1-g of Ferrous Sulfate [FeSO4.7H20] to the jar. This is a common component in fertilizer that Iron Bacteria can convert to Hematite.

With 24-hours incubation in my sample window, there showed a clear orange layer at the top of the jar [Sat. AM]. As of Tuesday evening, the clear orange layer had turned into rust red material floating at and just below the surface.

I have not run the test for Hematite yet [I will do so later], but I am fairly certain that it will turn out to be Hematite. I will also do a “count” of the rod bacteria to see if the population has substantially increased. The hypothesis predicts Hematite and an increase in the bacteria as they catalyze the Fe[++] to Fe[+++] reaction and gain energy.

I will add to this report as the data come in.

3 Responses to “Case of the Missing Lake Merced Hematite”

  1. nixie Says:

    so, does hematite react with HCl?
    goodluck with that as well!! :]

  2. John Hunt Says:

    Your HCl test is suggestive of an iron oxide. However, this type of test cannot tell you what type of iron oxide mineral you have (e.g., hematite). For that you need an XRD.

    I doubt you have hematite. Your sample is more likely ferrihydrite.

  3. John Hunt Says:

    Microorganisms don’t produce hematite. They produce ferrihydrite, which can recrystallize into hematite.