Interpreting field methods

You can improve your ability to design valid field studies in vegetation science by evaluating the work of others. For the following five studies, puzzle out the actual design. A few hints:

• Use the entire description of methods to determine the actual design.
• The actual sampling universe is the set of possible observations, the patch of nature to which inferences can legitimately be made. The actual sampling universe, as revealed by the sampling design, can differ from the sampling universe that the author might have had in mind.
• You can get a hint on the number of independent observations, in other words the number of replicates, by thinking about the calculations for mean and standard deviation. In the equation for standard deviation, n is the number of independent observations.
• Sometimes designs can be interpreted in different ways. In these examples there is only one completely correct and consistent answer.
• This exercise is more subtle that it appears.

Read the descriptions of field methods in each of the following five studies, then pick the best answer among the choices. Submit your multiple-choice answers in a report; no text or explanation is necessary.

Study A

"The study area was a 2 ha agricultural field at Hyslop Experimental Farm. For sampling, I selected at random 160 1-m² quadrats from a grid of 20,000 possible quadrats within the field. I measured the number of individuals of three weed species within each quadrat."

What is the actual sampling universe?

1. Hyslop Farm
2. All similar agricultural fields
3. The 20,000 quadrats mentioned in the methods
4. The 160 quadrats mentioned in the methods

1. 1
2. 3
3. 160
4. 480
5. 20,000

Is the sampling per area or per individual?

1. Per area
2. Per individual

Study B

"The study area was Watershed 13 at the H. J. Andrews Experimental Forest. All 4521 trees in the watershed were numbered from 1 to 4521. I selected 100 numbers at random from 1 to 4521 and located the trees with those numbers. I then measured the basal area and height of each of those trees."

What was the actual sampling universe?

1. The H. J. Andrews Experimental Forest
2. Watersheds similar to Watershed 13
3. The 4521 trees within Watershed 13
4. The 100 trees mentioned in the methods

What was the number of independent observations of tree basal area?

1. 1
2. 13
3. 100
4. 4521

Was the sampling per area or per individual?

1. Per area
2. Per individual

Study C

"I sampled within 2-m × 5-m quadrats with the long axis oriented parallel to the river edge. To pick sampling locations, I first determined the length, 1535 m, of the section of interest along the east bank of the Deschutes River, Oregon. I then picked 20 random numbers between 0 and 1530. I then put the random numbers in order. Starting at one end of the section, I moved the number of meters corresponding to the first number and placed the end of the quadrat at that point for sampling. I then moved to the location matched by the second number and placed the quadrat for sampling, and so forth until 20 quadrats were sampled along the bank. Within each quadrat I measured the cover (%) of five shrub species."

What was the actual sampling universe?

1. Rivers in central Oregon
2. The Deschutes River
3. The Deschutes River from the first quadrat to the last
4. The 2-m strip along the east bank of the Deschutes River from the first quadrat to the last
5. The 2-m strip along the east bank of the Deschutes River from the starting point (0 m) to the ending point (1535 m)

What was the number of independent observations per species?

1. 1
2. 5
3. 20
4. 100
5. 1535

Was the sampling per area or per individual?

1. Per area
2. Per individual

Study D

"The study area was a 2200 m by 2200 m forest tract in the Bull Run watershed. To pick sampling locations, I picked 15 pairs of random numbers between 0 and 2200, representing distances along each edge of the study area.  I used these numbers as Cartesian coordinates to locate 15 points.  At each of the 15 coordinate locations, I recorded the tree basal area of all trees within a 300-m² circular plot centered on the coordinates.  I then calculated stand basal area for the location by dividing the sum of the tree basal areas by 300 m².  I repeated this procedure at all 15 locations."

What was the actual sampling universe?

1. The Bull Run watershed
2. The forest tract mentioned in the Methods
3. The 15 plots
4. The trees that were measured

What was the number of independent observations?

1. 1
2. 15
3. 300
4. 2200
5. Can't tell without more information

Was the sampling per area or per individual?

1. Per area
2. Per individual

Study E

"The study area was a 10-ha unit of sagebrush scrub vegetation in John Day National Monument. At each of five random locations within the study area I placed a grid of nine 1-m² quadrats. Within each quadrat I measured the cover (%) of Agropyron spicatum and Bromus tectorum."

What was the actual sampling universe?

1. Sagebrush scrub vegetation
2. The John Day National Monument
3. Sagebrush scrub vegetation within The John Day National Monument
4. The 10-ha unit mentioned in the Methods
5. The five random locations

What was the number of independent observations per species?

1. 1
2. 5
3. 9
4. 18
5. 45
6. 90

Was the sampling per area or per individual?

1. Per area
2. Per individual

Version: May 31, 2008