Lac de Salanfe and Ottans 2015

From BioDesign for the Real-World
Jump to: navigation, search
Lake Ottans Valais.jpg


Collecting and testing water with the hermetic vial of arsenic bioreporter in the field for the first time.
We want to test if the water of the two lakes is contaminated with arsenic and also test our prototype in real conditions.
That is why we will make two sets of measurements :

  • Collecting water on the field and bringing it back to the lab for testing
  • Collecting and testing water samples directly in the field


Wednesday, August 12th

7:21 AM Platform 5, Lausanne Train Station


We want to collect water in the lake of Salanfe (1925m) and in the lake of Ottans (2171m), which are located above Martigny, in Valais/Wallis (Switzerland).

Arsenic Mine in Valais

These sites are particularly concerned by arsenic contamination because of the former exploitations of [arsenic mines] in the surrounding mountains. The arsenic and gold-mining stopped in 1928.
The mineral content and their references can be found in the public database organized by [here].

Suggested Routes

A Hikers path -

    • Take the public transport to Marécottes (1110m) (about 40 minutes from Martigny)
    • Bus to Van d'en Haut (1395m)
    • Hike to Auberge de Salanfe (1950m) (1h30)
    • Follow the path to Col d'Emaney (2462m) to arrive at the Lac des Ottans (1h45)
    • Lunch + water sample collection
    • Hike to Le Luisin (2785m) (0h45)
    • Hike down to La Creusaz (1805m) (0h30)
    • Take the telecabin to Marécotte
    • Marécotte to Lausanne via Martigny


    • Take the Bus to Marécotte telecabin
    • Take the telecabin to La Creuse
    • Hike to the peak of La Luisin (2785m)
    • Follow the path down to Col d'Emaney (2462m) to arrive at the Lac des Ottans

Then reverse the path back.


Here are the important points mapped out.

What we did

Due to timing, we did a loop

    • Take the public transport to Marécottes (1110m) (about 40 minutes from Martigny)
    • Bus to Van d'en Haut (1395m)
    • Hike to Auberge de Salanfe (1950m) (1h30)
    • Follow the path to Col d'Emaney (2462m) to arrive at the Lac des Ottans (1h45)
    • Lunch + water sample collection
    • Hike back to Van d'en Haut (1395m)
    • Bus to Marécottes (1110m)
    • Public transport from Martigny

In retrospect, it may be better to take a car, so not to be limited by the bus or telecabine schedules. Otherwise, it may also be nice to do this as a 2 day event.


1) From Lausanne to Martigny :

  • Departure : 7.21am - platform 5 (From Montreux : 7.42am - platform 3)
  • Train : InterRegio 1707 - Direction : Brig
  • Arrival : 8.12am - platform 2

2) From Martigny to Les Marecottes :

  • Departure : 8.43am - platform 40
  • Train : Regio 26210 - Direction : Vallorcine (F)
  • Arrival : 9.04am

3) From Les Marecottes gare to camping Van d'en Haut :

  • Departure : 9.14am
  • Bus : 12.225 : Salvan - Les Marecottes - Telecabine La Creusaz - Van d'en Haut
  • Arrival : 9.34am

Then we will continue by foot until La Creusaz via the Lac d'Ottans (see the proposal hiking). At La Creusaz, we will take the cable cars to come back to the valley.
4) From La Creusaz to Les Marecottes :

  • Departure : 1.30, 2.30, 3.30, 4.30 or 5pm
  • Telecabine des Marecottes
  • Arrival : 7 minutes later

5) From Les Marecottes telecabine to Les Marecottes gare :

  • Departure : 1.54, 2.54, 4.54 or 5.54pm
  • Bus : 12.225 : Salvan - Les Marecottes - Telecabine La Creusaz - Van d'en Haut
  • Arrival : 4 minutes later

Or by foot, 12 minutes via rue Principale and chemin de la Lenaire.

To come back to Lausanne, we have two choices :

6) From Les Marecottes to Martigny :

  • Departure : 3.12pm
  • Train : Regio 26215 - Direction : Martigny
  • Arrival : 3.35pm - platform 40

7) From Martigny to Lausanne :

  • Departure : 3.48pm - way 1
  • Train : InterRegio 1730 - Direction : Genève-Aéroport
  • Arrival : 4.40pm - platform 7

6) From Les Marecottes to Martigny :

  • Departure : 5.12pm
  • Train : Regio 26219 - Direction : Martigny
  • Arrival : 5.35pm - platform 40

7) From Martigny to Lausanne :

  • Departure : 5.48pm - way 1
  • Train : InterRegio 1734 - Direction : Genève-Aéroport
  • Arrival : 6.40pm - platform 7


Preparing for the Hike

    • Good hiking shoes
    • Hiking poles (optional, recommended)
    • Rain gear
    • Water / hydration - last time we went there, there was no drinking water from Van d'en Haut until the auberge
    • Your lunch and some food - keep it separate from the scientific materials!
    • Hat / sunscreen

Preparation of Acidified Samples for later laboratory determination of Arsenic by Bioreporter


Based on the report of the last Indian collection water trip, and the protocol for water collection.
we need to prepare some materials :

  • Recipient for collected samples : The bottles we will use must be imperatively in PET (glass absorb arsenic). Moreover we need a amount of sample of about 50mL.
    If there is iron in the water, it may be oxidized by atmospheric oxygen and precipitate, removing arsenic from the solution by sorption/co-precipitation. Therefore the pH of the sample must be brought down to below 2 [Trang et al., 2005]. To acidify the sample, we need to add to the sample 1mL of nitric acid (0.75M). Thus we need to take a bottle with nitric acid and a 1000uL pipette with tips (and gloves).
    We have to determine how many samples we want to collect.
We have taken 50mL plastic Falcon Tubes (centrifuge tubes). To wash them, we have washed with 5mL of 150mM nitric acid, and afterwards we have washed with water 3 times.
  • Deep water sampling device : We already have a solution to collect water on a lake. It is a jar, which allow sthe collection of water in depth. It is designed that the final sample does not touch the air in the device.
  • Ice pack : optional, but it is better to conserve samples at a temperature of 4°C.
  • Information card : To identify each samples, we have to collect some information about the site : GPS location, source of water, color/smell/clarity... we can note on a card. To perform it we have to prepare some cards.
  • GPS watch
  • transparent scotch
  • permanent pens for labeling
  • pH paper, thermometer… -> borrowed a special field tool kit for water collecting briefcase from ENAC with devices to measure O2, pH, temperature and conductivity
  • distilled water for washing the probes from the field tool kit
  • 150mM nitric acid solution for acidification
  • bubble pipette with 1ml gradation
  • gloves
  • waste bags

Methods on Site

First, we prepared the day before some 50mL tubes washed with nitric acid. We added some nitric acid in the tube and then we washed it with lab water.

Secondly, we followed this collecting water samples protocol.
To collect large amount of water, we use a special jar. After having blasted it with stones (it is important to collect water in deep), we bring it in the centre of the Lake. We just had to open the cap in the top of the jar to fill it until no bubbles went out of the recipient.
Then we collected some water from the second recipient with a 50mL syringe. We added a 0.8 um filter (Minisart) at the end of this syringe to filter the water. Finally we filled the previously acidified 50mL tubes with the samples.

After collecting water, we acidified it by adding 1mL of nitric acid to the 50mL of the sample for 3mM final nitric acid. We can analyse these samples during 2 weeks after acidification.
Non-acidified, non-filtered water was collected for later observation of any organisms.

Testing water in the field

NB : We are authorised to transport bacteria on the field by the BAFU (in vials with a silicon septum cap).


Before going on the field, we have to solve the problem of portability of the prototype by replacing power supply : batteries instead of a computer.
We also have to test the best way to transport active bacteria on the field.

  • Vials with bacteria prepared in MOPS working solution
  • ice pack in an insulated (styrofoam) box to keep the bacteria cold during hike
  • Syringe 50ml
  • Filter 0.8µm
  • Needle -> smaller gauge, short length, 1 per water sample + 1 for letting out air
  • Needle waste box -> a hard container to bring back used needles
  • Threshold standard -> 100x Concentrated arsenic stock to dilute (1000 ug/L) to make the threshold standard (10 ug/L, the WHO limit on drinking water) + to spike the water samples
  • Pipette 2-200µl size
  • 200ul Pipette tips -> 1x per sample + 1 for spiking the samples with concentrated arsenic stock
  • Waste bag
  • Tube rack
  • Gloves
  • Parafilm

We also need to bring our prototype. For the moment, it is in a wooden wine box. So we have to protect the box and fix the prototype on it. Thus we made a niche in polystyrene.
Moreover we protected the arduino and electronic part with a tips box, maintained by a elastic cord.

GMO Reporter Bacteria Preparation

Before being active, the bacteria must be incubated during about 2 hours at 37º and 180rpm. Then we must centrifuge the bacteria down, and resuspend the cells in the MOPS working solution. These operations take time (approximately 2.5h), before going hiking.
That is why, we tested whether the bacteria prepared and stored overnight at 4o can be used the next day.

The next morning performed the arsenic standard curve with laboratory tap water by incubating the bacteria in contact with different known concentrations of arsenic. After incubation, we measured the relative fluorescence with the fluorimeter and with our prototype after 6 h at 30oC and shaking.

* Storage of bacteria in MOPS working solution overnight in the fridge does not affect their activity.
* Prepare bacteria in MOPS working solution, aliquot in hermetic sample vials at 1.5ml each and store overnight at 4oC prior to hike
* Number of vials to prepare
** n=3 no bacteria (MOPS working buffer) for 100% transmittance
** n=3 with bacteria for the 10ug/L As(III) standard
** n=6 with bacteria per water collection site (n=3 as is, n=3 spiked to have 10ug/L final As(III))
100x Concentrated Arsenic Stock Preparation

To eliminate any excess air, 1000 ug/L of Ars(III) was prepared in laboratory distilled water in a screw-cap vial, then air in the vial was replaced by a gentle stream of argon gas. The cap was sealed with parafilm.

Methods on Site

First, we prepared the day before enough vials with bacteria. We put the bacteria in MOPS working solution and conserved them to the fridge at 4º (This method had been previously tested and approved).

Secondly, we followed this collecting water samples protocol.
To collect large amount of water, we use a special jar. After having blasted it with stones (it is important to collect water in deep), we bring it in the centre of the Lake. We just had to open the cap in the top of the jar to fill it until no bubbles went out of the recipient.
Then we collected some water from the second recipient with a 50mL syringe. We added a 0.8 um filter (Minisart) at the end of this syringe to filter the water. Finally we filled two 50mL vials with the samples.

One of the samples was spiked with 10ug/L arsenic (from 100x concentrated arsenic stock), the other one was not.

For each samples we collected :

  • In the Ottans Lake, in deep (about 10m)
  • In the waterfall between the Lake and the cascade enter

and for water from the Auberge de Salanfe (used as lab water), we filled 3 vials with the spiked sample and 3 without.

The vials were bring back in the lab at room temperature (close to 30º), but not in a incubator.
After 6h of incubation, we measured the turbidity (during 10s) and the fluorescence (during 30s) of the samples with the prototype 1.


Characteristics of the water we collected

During our water collection trip, we collected some samples in the Ottans Lake, which must be contaminated because of former arsenic and gold mining exploitation.

For each samples we measured the pH, the temperature, the conductivity and the dissolved oxygen.
Moreover we take some information about their GPS location and the water quality.

We took samples at 3 different point : • In the edge of the Lake (N 46º12'9590 E 06º95'7607") • In deep, in the middle of the Lake (N 46º12'9607 E 06º95'7298) • In the waterfall which is between the mine enter and the Lake (N 46º12'9688 E 06º95'8167").

Water quality

The water, after filtration, is clear, without smell and colour.


Figure 1 : pH measured for the 3 samples

Temperature in degree

Figure 2 : Temperature in degree measured for the 3 samples.

Conductivity in uS/cm

Indication of salt presence in our samples?
Figure 3 : Conductivity measured in the 3 samples.

Dissolved Oxygen (DO) in percentage

Reference DO of the air : 98%
Indication of the movement of the water in the Lake?
Figure 4 : Dissolved oxygen measured in the 3 samples.

Water test in the field

The test we wanted first to do with the samples we collected was : measured the turbidity and the fluorescence thanks to our prototype 1 after 6h of incubation. With these measure we also can calculate the absorbance with the formula : 1+LOG(turbidity/mean fluorescence of the buffer,10) and the relative fluorescence which is the fluorescence/absorbance.
To perform this measurement, we prepare some bacteria the day before, by putting them into MOPS working solution and keeping them into the fridge at 4º during the night. This method had been tested before and gives usable results.
After having filtered the samples we collected using a 0.8um filter (Minisart), we put them in the vials containing the bioreporter. The measures have been taken after 6h of incubation at atmosphere temperature.

In this case we analyzed 3 samples : from the middle of the Ottans Lake (in deep), from the waterfall and water from the Auberge de Salanfe (as lab water).
For each samples, we made the measurements with the samples and with the samples spiked with arsenic at 10ug/L as a control.

Figure 5 : Relative fluorescence (Fluorescence/Absorbance measured with the prototype 1) and standard deviation for the three samples with (10ug/L added, in red) and without spiking (in blue).

Positive control - Water with and without spiking :

  • We get the results we expected higher relative fluorescence in the samples where we added 10ug/L arsenic than this one without spiking.

Waterfall samples :

  • Low error bar : significant measures

Lake samples :

  • High error bar : make more measure to have significant results

Some error bars are big because of one measure, which is clearly different from the other one. This different could come from different parameters :

  • Problem of counts because of the prototype 1
  • More iron or particles in this sample
  • Samples toxicity for bacteria

In the case the measures are not satisfying, that is why we will complete them with the water analysis in the lab, with the samples we bring back.

Then we calculate the ratio of spiked vs not spiked samples.
Figure 6 : Ratio spiked/not-spiked samples for the three collect location.

We expected to have the same ratio for all the samples, sign that the measures appear to be in the same conditions.
In our case, water has a ratio around 4 and the two other samples have a ratio close to 2 (2 time lower than the water ratio).
So something must have disrupted our measurements of the lake and the waterfall samples, for example iron or particles in the samples which is able to diminish the response of the bioreporter.

We are also interest by comparing the calculate absorbance of the samples. Absorbance increases with the presence particles in the samples, like bacteria, arsenic, iron…

Figure 7 : Absorbance (1+LOG(turbidity/mean fluorescence of the buffer,10)) for the three samples with (10ug/L added, in purple) and without spiking (in green).

We note first that we have a big error bars for the same samples in the figure 5. So maybe the problem comes from the turbidity measurement with the prototype 1 or from the presence of more and bigger particles in these samples.
Moreover we expected to have more absorbance in samples with spiking. In our case, we see that for water and the samples from the middle of the Lake we have an absorbance, which is a little bit lower. But it is also for these measurements that the error bar is big, so maybe they can be wrong.

We finally looked at the fluorescence emit by our samples. Fluorescence comes from the eGFP protein couple with the transporter that bacteria produce in presence of arsenic.

Figure 8 : Fluorescence measure with the prototype 1 for the three samples with (10ug/L added, in blue) and without spiking (in yellow) and their standard deviation.

As expected we have a clear higher fluorescence for the samples with spiking than without. In all the case, the error bars are not important, so the fluorescence measure with the prototype 1 seems to be satisfying.

Water test in the lab

Neutralization of the acidified samples

To recall, the samples which has been collected on the field, has been acidified and fridge at 4° for their conservation.

When we need to use a sample, we just have to neutralize the acid with the help of a pyrophosphate solution.

In our case the neutralization has well worked. As we see in this picture before the neutralization, the pH is around 4 (at left) and after it is at 7 (at right).

1st test : no conclusive results

After neutralization, we tested five samples from :

  • The Ottans lake, in deep, from the little recipient of the jar
  • The Ottans lake, in deep, from the big recipient of the jar
  • The waterfall, between the mine enter and the lake
  • The Auberge de Salanfe and the lab, as positive control.

In each case, we have one sample skiped at a arsenic final concentration of 50ug/L and one without spiking.

In a 96-wells plate, we loaded 100uL of bioreporter solution and 100uL of samples with and without spiking. We let them incubate at 37° in a rotary shaker and we took a measurement of fluorescence and absorbance after 2, 4 and6 hour of incubation.

We calculated the fluorescence/absorbance for each samples and at each time and we plotted them :
Fluorescence/absorbance for the five samples after 2, 4 and6 hour of incubation.

The results we get at 2 and 6 hours are not workable. Indeed the positive controls show a bigger or similar fluorescence/absorbance for the sample without spiking than with spiking. However we expected to have a higher fluorescence for the sample with spiking.
The problem may come from the amount of arsenic we add when we spiked which could be too big. In this case, the bacteria can saturate.

The results at 4 hour are more reasonable, because they show a bigger fluorescence/absorbance for the spiked samples. So the positive control works.
We also have a higher fluorescence for the lake samples, sign of a bigger concentration of arsenic in the Lake, than in the waterfall or in the positive control water.

As these results are not conclusive, we have to make other experiments and adding these modifications :

  • Spiking with different concetrations,
  • Measuring the fluorescence with the same conditions to be able to compare the results at different times.
2nd test : calculation of the arsenic concentration in the samples

The goal of this second test is to detect the presence of Arsenic on our samples and to calculate its concentration.

Anew, we measured the fluorescence and absorbance after 2, 4 and 6 hours of incubation thanks in the following samples (after neutralization) :

  • Ottans lake water, in deep, from the little recipient of the jar
  • Ottans lake water, in deep, from the big recipient of the jar
  • Waterfall water, between the mine enter and the lake
  • Auberge de Salanfe and the lab water, as positive controls.

For each samples, we perform three spiking at 10, 30 and 50 ug/L of Arsenic final concentration and one without spiking.

Then we apply a method used in Pr. Van der Meer lab to calculate Arsenic concentration in urine, explain in the following picture :

The principle is to plot the fluorescence/absorbance in function of the final concentration of Arsenic with and without spiking.
With these points (at least 2, but more precise with more points) we can plot the regression line with Excel.
We get some values :

  • a' : the slope of the reference curve, here we take water from the lab as reference
  • a : the slope of the sample in which we want to measure the Arsenic concentration
  • b : the y-intercept of this sample
  • y-sample : the value of fluorescence/absorbance without spiking

And from these values, we can calculate x = (y_sample - b')/a' where x is the Arsenic concentration we are looking for and b' = a'/a * b.

In our case we find the following concentration values for our samples after 2, 4 and 6 hours of incubation :

Sample location [As] after 2h incubation [As] after 4h incubation [As] after 6h incubation
Lake water collected in deep, from the little recipient 2.41 ug/L 13.12 ug/L 17.52 ug/L
Lake water collected in deep, from the big recipient 0.69 ug/L 7.16 ug/L 12.33 ug/L
Waterfall water 7.47 ug/L 4.11 ug/L 0.89 ug/L
Auberge de Salanfe water 5.50 ug/L 1.71 ug/L 1.80 ug/L

According to this paper, we expected to find an Arsenic concentration of about 150 ug/L in the Ottans lake. In our case, the results are 10 time smaller.
We can suppose that the samples at 2h need more time of incubation to give workable results. That is why the results we get at 6h are more probative.

After having recieved the feedback from Pr. Van der Meer about our results, it appears we made some mistakes which can explain our wrong values :

  • We have to spike at lower final concentration of Arsenic than 30 or 50 ug/L (10 ug/L is more corrected), because in this case the Arsenic concentration in our samples is bigger than 100 ug/L and our bacteria are saturated.
  • Errors in the regression curve can be increased by error of pipeting. For example, it is better not to change the pipete with which we add Arsenic (spiking) because they can be calibrated differently.

In conclusion, these results prove that we are able to detect Arsenic in samples collected in the field, but we have to be more precise in the manipulations.


Feedback for Future Trips

  • Syringes are not ideal in the field

Other minor comments

  • Make collapsible tube racks - they are not a luxury but a necessity
  • Put depth marks on the ropes of the deep water collector
  • Manage the ropes better on the deep water collector (maybe a system similar to fishing rod)


Pfeifer H.R., Häussermann A., Lavanchy J.C., Halter W., 2007. Distribution and behavior of arsenic in soils and waters in the vicinity of the gold-arsenic mine of Salanfe, Western Switzerland. Journal of Geochemical Exploration 93 pp. 121-134. 10.1016/j.gexplo.2007.01.001 [Journal Site]