Exercise 4:  Analysis of stream habitat unit data

In this exercise, you will analyze a set of stream inventory data, and compare channel morphology and the abundance and qualities of channel (habitat) units in two reaches. You will work with a set of data for Hehe Creek in the Fall Creek watershed of the Willamette National Forest. Data for two reaches, 3 and 6, are available. The locations of these two reaches are shown on the map (under construction).

Field Procedure: In each reach, habitat (channel) units were identified in the field. Each habitat unit is identified by with a Natural Sequence Order number, with NSO 1 being the first habitat unit at the mouth of Fall Creek, and NSO numbers increasing upstream from that point. Reach 3 consists of NSO 107-145; NSO 107 is the downstream-most unit. Reach 6 consists of NSO 316 to 368; NSO 316 is the downstream-most unit. Several characteristics (variables) are recorded for each unit, including dimensions, substrate (bed material), LWD, and fish cover. The length and width of each habitat unit were visually estimated in the field. A subsample of habitat units was also measured with a tape. By comparing the estimated HU length with the measured HU length for the subsample, a correction factor was determined. The length and width values in X4data are corrected values. Depth was measured at each unit, so no correction is needed for depth. All dimensions are in feet and tenths of feet (not inches).

Variables in the data file:

L: length of unit going upstream.

W: width of the water surface of the unit at the time of measurement; not bankfull width but a low flow width; sometimes called wetted width.

D or max. D: maximum water depth in the unit (at the time of measurement, not bankfull).

D ptc: measured at pools only; depth at the pool tail crest, which is the sill separating the pool from the next unit downstream.

Substrate Dom.: size class of the dominant (most abundant) particles on the wetted stream bed.

total cover -- % class: percentage of the unit that provides fish cover; percentage is not measured to nearest percent but one of four percentage classes is selected.

total cover – dom.: dominant type of cover; what creates the cover.

Most of the variables – unit dimensions, # pieces of LWD – are continuous variables. For these variables, the values represent the exact distance or difference between observations; it is valid to calculate averages, standard deviations, etc. The substrate variables and the cover variables are categorical variables; the value is not a real number but a symbol for a class. It is not valid to calculate averages for these values.

Part A: Overall analysis by reach

1. Do this part on the sheets with units in natural sequence order. Do one reach first, then repeat the steps for the other reach.

2. Do some calculations to get average values for the reach. Do these calculations (for the units making up the main channel only, not including side channel, tributaries and the culvert) in the empty rows below the unit data. In the first empty cell in the L column, calculate average L for the reach. Do this by inserting the function =average(range) in this cell. You can use the command Insert – Function to find and paste functions in a cell. The range should be all the cells containing unit values for L (of mainstem units – not including T and S units). Be sure that you include all of the appropriate cells in the formula. Then calculate average W, and average max. D for the reach. Do these calculations in the first empty cell below the respective columns. Now you have a series of average values in this row; label the row Average in the left-most cell. Then calculate average L, average W, and average max. D for the other reach.

3. Now compare substrate size between the two reaches. Hydraulic geometry theory suggests that substrate size decreases going downstream, but many mountain streams with bedrock influence and constrained valleys don’t follow this rule. You can’t calculate average substrate size for each of the two reaches, since we have only categorical data for substrate size. An alternative is to determine the mode of dominant substrate and subdominant substrate for each reach. Go to the next empty row below the row of average values (insert a new row if needed), and label it Mode (in the left-most cell). In the cell of this row that falls below the column containing Substrate – Dominant, enter the function =mode(range). The range should be all the cells containing unit values for Substrate – Dominant (of mainstem units – not including T and S units). The value returned from this function is the most frequently occurring size class for dominant substrate. Do this for subdominant substrate. Find the mode for dominant substrate and subdominant substrate for the other reach.

4. Now determine amount of fish cover for the reach. Use the mode function to calculate the mode of % cover classes. Do this for the other reach.

5. Calculate LWD loading for the reach. Sum the total number of pieces of LWD for the reach. Determine reach length in miles, in the following way. Sum the L values to get total length of the reach, in feet. Divide total reach length by 5,280 to convert to miles. Divide total number of pieces of LWD (all 3 classes) by total reach length in miles to get LWD loading, in pieces per mile.

6. For all formulas, check to make sure that you have included all the relevant cells in the range, and no other cells.

7. Repeat the analysis for the other reach.

8. Make a table summarizing the average and modal characteristics of the two reaches. Include all the characteristics you have calculated in steps 2-5. Include this table in your answers to the A questions.

9. Print out your completed spreadsheets for Part A and include them with your answers

Questions for Part A: Answer each question briefly.
A-1. How are Reaches 3 and 6 different in terms of their width and depth? Do these results fit hydraulic geometry theory? If not, how not?

A-2. How are Reaches 3 and 6 different in terms of substrate?

A-3. How are Reaches 3 and 6 different in terms of LWD loading and % fish cover?

A-4. Do the modal values for substrate and % fish cover accurately represent the data, in your opinion? Scan the data for the individual units to see if the modal value is representative.

Extra credit charts and question (2 extra points on 10 point exercise) :
It may be more meaningful to look at the distribution of categorical values, rather than simply the mode of the categorical values. The distribution can be displayed in a histogram. Make a histogram of Substrate – Dominant for Reach 3, and another one for Reach 6. Then make a histogram of % cover class for each reach. You can make a histogram using the command Tools – Data Analysis – Histogram. You will need to start by setting up a short column of values to be used as the bins for the histogram. Do this in a column to the right of your data. Put the values 1, 2, 3, 4, 5, 6 sequentially in the column. Then specify this range for the bins in making the histogram. If you can’t get this to work, make your histograms by hand.
A-5. What do the histograms show you about these two variables (Substrate Dominant and % fish cover) for each reach that the modal values did not show you? (Include your histograms with your answer.)

Part B: Analysis by unit type

1. Do this part on sheets with units sorted by type. You will probably want to do the analysis for one reach first, then repeat the analysis for the other reach.

2. In a new column, calculate the area of each unit in the reach, by multiplying L by W. Be sure to label the new column.

3. In a new column, calculate residual pool depth for each pool in the reach. Residual pool depth equals maximum pool depth minus Dptc. Residual pool depth is considered an important indicator of fish habitat quality; large residual pool depth indicates good habitat.

4. Now do some calculations for each unit type in this reach. Do these calculations in the empty rows below the pools, below the riffles, etc. You may have to insert a few more rows. In the first empty row below the pools, calculate the average L of pools in the reach. Do this by inserting the function =average(range) in the empty cell below the column of numbers on which the average will be calculated. Be sure that you include all of the appropriate cells in the formula. Then calculate average W, average max. D, and average residual pool depth for the pools. Then calculate average L, average W, and average max. D for riffles. Do this separately for the pool and riffle units in the reach. (There is only one glide and one fall, so averages are not needed.)

5. In the empty row below the averages, calculate the sum of pool area for the reach. The sum is needed for the areas only, not for the other dimensions. Do this for the riffles also.

6. At the bottom of the data cells, use some empty cells to first calculate the total area of all fall, glide, pool and riffle units for the reach. Use the function =sum(range) for this. Then calculate the % of pool area in this total unit area for the reach. % pool area is equal to 100*(sum of pool areas)/(total area of the reach). Do the same for % riffle area, % fall area, and % glide area.

7. It is generally accepted that pools have finer substrate than riffles. To test this, determine the mode of dominant substrate and subdominant substrate for each unit type. At the bottom of the pools, create a new empty row and label it mode. In the column containing Substrate – Dominant, enter the function =mode(range). Do this for subdominant substrate. Find the mode for dominant substrate and subdominant substrate for riffles. Then repeat this for the other reach.

8. We would also expect pools to have better cover for fish than riffles. Use the mode function to calculate the mode of % cover classes, for pools and for riffles. Do this for both Reach 3 and Reach 6.

9. For all formulas, check to make sure that you have included all the relevant cells in the range, and no other cells.

10. Repeat the analysis for the other reach.

11. Make a table summarizing unit characteristics of the two reaches. Include all the unit characteristics you have calculated in steps 1-9. Include this table in your answers to the B questions.

12. Print out your completed spreadsheets for Part B and include them with your answers.

Questions for part B:
B-1. How are pools of Reach 3 different from riffles of Reach 3, in length and depth? How are pools of reach 6 different from riffles of reach 6, in length and depth?

B-2. Which reach has more pool habitat? That is, how does pool abundance (pools as a % of total reach area) compare between the two reaches? How does pool size compare between the two reaches? How are riffles similar or different between the two reaches?

B-3. Do pools have finer substrate than riffles in Reach 3? In Reach 6?