Choosing Sites
Both the Moormans' and Mechums' watersheds were copied onto tracing paper, along with each river's order (1,2, or 3). The class divided into two groups and independently devised their own methods of site selection.
The Mechums Group divided the Mechums' watershed into four distinct areas whose bounds ran east to west enclosing distinct branches of each watershed; the goal was to ensure the balanced distribution of orders across the four areas. After this division, each river order was broken up into sections bound by the order's convergence with other river orders and then all sections were numbered within their given area and order; a calculator was used to randomly select one section of an order 1 and one section of an order 2 from each distinct area. (The river with the number corresponding to that of the one randomly generated by the calculator was selected for testing.) Two of the four areas were chosen to select two order threes in addition to the order 1's and the order 2's. The random selection process was repeated and one section of an order 3 was selected from each of the two previously selected areas. In total, the Mechums Group tested four order 1's, four order 2's, and two order 3's.
The Moormans Group randomly selected sites within four distinct areas running North to South. The Moormans group randomly selected an order one and an order two from each area (by the random number generation method explained above), but did not break these orders into specific sections; locations for testing were chosen by convenience relative to roads. One order three was selected by the same method as that of the Mechums Group, and finally, the only order four found in both watersheds was also tested. In total, the Moormans Group tested four order 1's, four order 2's, one order three, and one order four.
Site Protocol
At each river, we measured DO, BOD, turbidity, conductivity, temperature, phosphate, nitrate, flow rate, pH and E. coli. We split the river into sections of near bank, middle of the river, and far bank. At rivers that were not very wide we would probe the length by above, middle, and below. All of these parameters, except E. coli, were measured using Vernier probes and Lab Quests. For E. coli, we used a Coliscan water testing kit.
Measuring Nitrate and Ammonia
To measure Nitrate and Ammonia, the probes have to be calibrated in solutions for 30 minutes. Originally, we brought the probes out with us to the field, but as the year progressed we started the calibration in the lab and then collected the water samples at the river, and brought the water samples back to the lab for testing. We concluded that the data for Ammonia or Nitrate would still be the same if we measured it at the river or back in the lab because we collected data and ran an experiment. After the experiment, we decided that there was not enough evidence to prove that delaying measurement mattered or affected the final data.
Measuring DO, Turbidity, pH, Temperature, Conductivity
These values were measured in the field using Vernier probes. Samples were brought to the steam-side table and measured on Lab Quests. For conductivity we had the Vernier probe set at 0-200 micro siemens, in the future we should have it set at 0-1500 micro siemens.
Flow Rate
The Flow of the river was measured using a Vernier Flow Rate Sensor. The sensor was placed at a 90 degree angle at the bottom of the river at each of the six sites where the water was previously taken and a reading of the river velocity was recorded. According to protocol we should have measured the depth of the river to calculate the full discharge of the river however we did not follow this protocol. In following years it would be better to follow this protocol to obtain better readings.
BOD
A sample was collected in a water bottle covered with aluminum foil (to prevent light from entering sample). The sample was allowed to sit in the lab for at least 24 hours. At a convenient time the sample was tested for DO using a Vernier probeThis value was subtracted from the DO value recorded at the site on the day of collection (initial). The difference in DO is the DO used or BOD. This value was divided by the hours in which the sample sat in the lab covered with aluminum foil. The resulting %saturation/hr was recorded in the data chart. In the future it would be best to convert this to mlDO/hr.
Measuring Phosphate
No probes are available for the measurement of phosphate ions. As a result we used a colorimeter (Vernier) to measure transmittance of light through a Phos Ver 3 Phosphate reagent.
Procedure:
1. Collect 10 ml of sample water
2. Add one 10 ml sample of Phosphate Reagent (Phos Ver 3 Phosphate Reagents (Hach :Cat # 21060-69)) to the 25 ml sample.
3. Swirl solution until the powder is dissolved. (about 5 minutes)
4. Fill plastic cuvette included in the Vernier Colorimeter test kit with treated water.
5. Calibrate colorimeter to 565 nm wavelength.
6. Follow directions for Vernier colorimeter reading .
7. Record transmittance value. The conversion to g/l was made. Three standard potassium phosphate solutions (0.01, 0.01, 0.001) were prepared in the lab. These three knowns were then tested for % transmittance using the same colorimeter used in the field. A regression line based on this set of 3 values was calculated. The following formula was derived (% transmittance-1.031)/9.466)/1000= xg/ml . The Phosphate values recorded in the data charts were derived from this formula.
Benthic
For two minutes we held a net in a rapid in a section of the river and agitated the bottom of the river. This released various benthic organisms from under rocks and from the bottom of the river, giving us a benthic data sample. After collecting the benthic organisms in the net, we would bring it to a table set up on the side of the river, and we would identify and record all of the organisms, until we reached 200. If we did not reach 200 organisms, we would collect a second benthic sample.
Data Analysis
After collecting all of our data, we used a statistical program called Fathom to organize data, make graphs,run Monte Carlo simulations, and find 90% confidence intervals.
For each of the following parameters followed the same steps:
Both the Moormans' and Mechums' watersheds were copied onto tracing paper, along with each river's order (1,2, or 3). The class divided into two groups and independently devised their own methods of site selection.
The Mechums Group divided the Mechums' watershed into four distinct areas whose bounds ran east to west enclosing distinct branches of each watershed; the goal was to ensure the balanced distribution of orders across the four areas. After this division, each river order was broken up into sections bound by the order's convergence with other river orders and then all sections were numbered within their given area and order; a calculator was used to randomly select one section of an order 1 and one section of an order 2 from each distinct area. (The river with the number corresponding to that of the one randomly generated by the calculator was selected for testing.) Two of the four areas were chosen to select two order threes in addition to the order 1's and the order 2's. The random selection process was repeated and one section of an order 3 was selected from each of the two previously selected areas. In total, the Mechums Group tested four order 1's, four order 2's, and two order 3's.
The Moormans Group randomly selected sites within four distinct areas running North to South. The Moormans group randomly selected an order one and an order two from each area (by the random number generation method explained above), but did not break these orders into specific sections; locations for testing were chosen by convenience relative to roads. One order three was selected by the same method as that of the Mechums Group, and finally, the only order four found in both watersheds was also tested. In total, the Moormans Group tested four order 1's, four order 2's, one order three, and one order four.
Site Protocol
At each river, we measured DO, BOD, turbidity, conductivity, temperature, phosphate, nitrate, flow rate, pH and E. coli. We split the river into sections of near bank, middle of the river, and far bank. At rivers that were not very wide we would probe the length by above, middle, and below. All of these parameters, except E. coli, were measured using Vernier probes and Lab Quests. For E. coli, we used a Coliscan water testing kit.
Measuring Nitrate and Ammonia
To measure Nitrate and Ammonia, the probes have to be calibrated in solutions for 30 minutes. Originally, we brought the probes out with us to the field, but as the year progressed we started the calibration in the lab and then collected the water samples at the river, and brought the water samples back to the lab for testing. We concluded that the data for Ammonia or Nitrate would still be the same if we measured it at the river or back in the lab because we collected data and ran an experiment. After the experiment, we decided that there was not enough evidence to prove that delaying measurement mattered or affected the final data.
Measuring DO, Turbidity, pH, Temperature, Conductivity
These values were measured in the field using Vernier probes. Samples were brought to the steam-side table and measured on Lab Quests. For conductivity we had the Vernier probe set at 0-200 micro siemens, in the future we should have it set at 0-1500 micro siemens.
Flow Rate
The Flow of the river was measured using a Vernier Flow Rate Sensor. The sensor was placed at a 90 degree angle at the bottom of the river at each of the six sites where the water was previously taken and a reading of the river velocity was recorded. According to protocol we should have measured the depth of the river to calculate the full discharge of the river however we did not follow this protocol. In following years it would be better to follow this protocol to obtain better readings.
BOD
A sample was collected in a water bottle covered with aluminum foil (to prevent light from entering sample). The sample was allowed to sit in the lab for at least 24 hours. At a convenient time the sample was tested for DO using a Vernier probeThis value was subtracted from the DO value recorded at the site on the day of collection (initial). The difference in DO is the DO used or BOD. This value was divided by the hours in which the sample sat in the lab covered with aluminum foil. The resulting %saturation/hr was recorded in the data chart. In the future it would be best to convert this to mlDO/hr.
Measuring Phosphate
No probes are available for the measurement of phosphate ions. As a result we used a colorimeter (Vernier) to measure transmittance of light through a Phos Ver 3 Phosphate reagent.
Procedure:
1. Collect 10 ml of sample water
2. Add one 10 ml sample of Phosphate Reagent (Phos Ver 3 Phosphate Reagents (Hach :Cat # 21060-69)) to the 25 ml sample.
3. Swirl solution until the powder is dissolved. (about 5 minutes)
4. Fill plastic cuvette included in the Vernier Colorimeter test kit with treated water.
5. Calibrate colorimeter to 565 nm wavelength.
6. Follow directions for Vernier colorimeter reading .
7. Record transmittance value. The conversion to g/l was made. Three standard potassium phosphate solutions (0.01, 0.01, 0.001) were prepared in the lab. These three knowns were then tested for % transmittance using the same colorimeter used in the field. A regression line based on this set of 3 values was calculated. The following formula was derived (% transmittance-1.031)/9.466)/1000= xg/ml . The Phosphate values recorded in the data charts were derived from this formula.
Benthic
For two minutes we held a net in a rapid in a section of the river and agitated the bottom of the river. This released various benthic organisms from under rocks and from the bottom of the river, giving us a benthic data sample. After collecting the benthic organisms in the net, we would bring it to a table set up on the side of the river, and we would identify and record all of the organisms, until we reached 200. If we did not reach 200 organisms, we would collect a second benthic sample.
Data Analysis
After collecting all of our data, we used a statistical program called Fathom to organize data, make graphs,run Monte Carlo simulations, and find 90% confidence intervals.
For each of the following parameters followed the same steps:
- Made a boxplot of the data and took out any outliers from the data and made a new, actual data set.
- Ran a Monte Carlo Simulation.
- Calculated a P value. Interpreted as to whether or not the parameter was different in each watershed.
- Constructed 90% Confidence Intervals on parameters where inference conditions applied and did not run confidence intervals where standards were not met.