3. How to measure: Measuring vegetation characteristics per area

Measuring vegetation in quadrats

Measuring biomass per area by the harvest method

Measuring density and frequency

Measuring cover in quadrats

Making and using quadrats

Line and point intercept methods of measuring cover

A final note on overlapping plants


Measuring vegetation in quadrats

To make measurements per area, you must somehow delineate areas (with just a few exceptions). The simplest way to delineate an area is with a quadrat frame. You then take your measurements within this quadrat. Measurements from multiple quadrats properly distributed through the community will allow you to extrapolate from your quadrats measurements to the whole community.
Quadrat

Measuring biomass per area by the harvest method

Although simple in principle, biomass measurements are tricky in practice. The basic technique for measuring biomass is the harvest method: Simply remove biomass, dry it in an oven to remove water, and weigh it. (Because the water content of plants can vary dramatically day to day, even hour to hour, drying to get "dry weight" is a more stable measure of biomass than is fresh weight.) An important choice is whether to measure total biomass (aboveground and belowground) or just aboveground biomass. Most biomass studies in vegetation science record aboveground biomass, for the straightforward reason that belowground biomass is very difficult to measure.

Let's start with how to harvest aboveground biomass. The first question is where the ground surface is! Some plants, like bunch grasses, have their highest biomass concentrations near the soil surface, so decisions about where to delimit aboveground from belowground can be very important. A good choice is to measure biomass down to the top of the soil profile. That is, measure all biomass above the soil, even if it is buried in litter. The best ways to harvest plants for aboveground biomass measurements vary with the growth form of the species. A serrated knife, like a bread knife, works very well for rooted, herbaceous plants. Shrubs require something heavier, like a garden loper or saw. When sawing shrubs for aboveground biomass harvest, try to cut as close to the soil surface as possible. Harvesting mature trees requires forestry skills, which are beyond the scope of this course.

When measuring biomass per area, harvest all plant material within the boundaries of the quadrat. That is, harvest the parts of plants within the quadrat, even if the plant is rooted outside the quadrat. Similarly, do not harvest parts of plants outside the quadrats, even if the plant is rooted within the quadrat. Other folks harvest the biomass of anything rooted within the quadrat, but I find it troubling, for several reasons. (a) Some species trail great distances, and it is hard to collect all of their pieces. (b) Many species are rooted in more than one location, making the "rooting" rule inconsistent. (c) The measurements are supposedly "per area" but because some species are large and more likely to extend beyond the quadrat, there is a confounding of area and individual.

 

Put the harvested material in a carefully labeled bag. The label should include the quadrat identifier, the collection date, and your name. If you are measuring biomass separately for each species, use a separate bag with the species's name on it. Likewise, if you are measuring biomass by plant parts (like stems, leaves, and reproductive structures), have separate, labeled bags for each. If you use grocery-quality paper bags, make sure the flaps are fully glued and won't leak pieces of your hard-won collection. Paper bags are better than plastic bags because paper is less likely to tear and paper bags can go directly into the drying oven. Metal collection tins also work.

Harvesting plants can be messy. Be sure to save any parts, like leaves, that become dislodged during the process.

For measuring dry-weight biomass, dry your samples in an oven hot enough to evaporate the water but not hot enough to break down the plant. Common drying temperatures are 65 C and 100 C. Dry your samples until all the water is evaporated. All the water is evaporated when the weight of the sample no longer changes. Small samples of herbaceous plants will dry to constant weight overnight. Large samples or woody material will take longer.

Bags from the oven can hydrate rather quickly, causing the weight you measure to be too high. Either place the dried bags in a desiccator, or weigh each bag within a minute or two of removing it from the oven.

After weighing the sample, you must account for the weight of the bag. The easiest way to do this is to weigh each bag before placing the biomass sample in it, making sure that the bag is completely dry. You might even want to put your collection bags in the oven before weighing them. If you weigh bags after drying, be careful to remove all the plant material and re-dry the bag. Weighing bags before collection has the advantage of allowing you to keep all your samples in their labeled bags, in case there is a mix up.

Measuring belowground biomass is difficult. One technique is to leave the plant in place and remove the soil around it. This small-scale version of placer mining is most feasible along road cuts or other places where the soil has already been cut away. Another technique is to remove the plant and its surrounding soil from the whole quadrat. Back in the laboratory, you can then carefully wash the soil away from the plant roots. Be careful to use gentle streams of water to wash away the soil so you minimize the loss of fine roots. It is almost impossible to harvest root hairs, because they bind so well to soil particles.

A machine called a hydropneumatic elutriator provides a mechanized way to separate organic material from inorganic soil (Smucker et al., 1982).

Another difficulty with harvesting belowground biomass is that roots are hard to identify to species. As a result, most studies of belowground biomass focus on the biomass of the entire community and not the biomass of individual species. Likewise, productivity and nutrient turnover rates are almost exclusively studied in terms of ecosystems. (An important exception is the measurement of productivity of agricultural crops (including forest plantations), where monocultures make ecosystem measurements and species measurements nearly synonymous.)

Special techniques are required to harvest bryophytes (like mosses and lichens). See McCune (1990) for details.

Measuring density and frequency

If you think that measuring biomass from field quadrats is difficult and time-consuming, you are correct. That is why vegetation scientists have developed other measures of plant abundance.

Density is the number of individuals per area. You measure density within a quadrat by counting the number of individuals and dividing by the quadrat's area. Density can be measured for all species or separated into the density of individual species or species groups. Usually, individual plants are counted only if they are rooted within a quadrat.

As mentioned in the Ecological Background chapter, density has little meaning when individuals of a species can vary greatly in size. Density is little used in vegetation science.

Frequency is the proportion of quadrats in which a species is present. Usually, a species is counted as present if a plant of that species occurs anywhere within the quadrat, whether or not it is rooted within the quadrat. Frequency values can vary from 0% to 100%. Frequency reflects both a species's abundance and how much it is spread over a community.

That Gleasonian bedrock, the Ecological Background chapter, shows has frequency also depends on quadrat size. As a result, frequency cannot be compared from community to community, from study to study, or from year to year unless the quadrat size is the same.

A note about distance measures for estimating density. Several measurement techniques use plant-to-plant or plant-to-random-point distances to estimate plant density. The basis for these techniques is that when plants are more crowded, they are closer to each other. Unfortunately, nearly all of these techniques assume that plants are randomly distributed across the study area. Since this assumption is seldom warranted, estimating plant density from distance measures is unreliable.

Measuring cover in quadrats

The most common measure of plant abundance is cover. Remember that cover is the proportion of the ground obscured by a species's aboveground leaves and stems (and flowers). (Cover can also be measured for the whole community by not categorizing plants into species.) Cover is popular because it can be measured quickly yet it reflects a plant's structural importance.

A cranky pronouncement about "percent cover"
Cover is often called "percent cover," as in "I measured percent cover." Cover is the attribute, but percent is the unit. Would you say "meter length"? No! So abandon "percent cover" and just say "cover."

There are several approaches to measuring cover in quadrats. One is the photographic method. When you want to measure cover that is essentially in a single stratum, with little overlap in cover between individuals, you can measure cover by taking a photograph and digitally calculating cover. This technique has been used mostly for measuring the cover of forest canopies, using fish-eye photographs. Recently a digital camera device was developed for agricultural settings, or when you do not need to distinguish one species from another, but it has already been discontinued. The photographic method has also been used with intertidal communities. But when the cover of plants overlaps each other, the photographic method won't work.

The most common way to measure cover is the visual estimation method. Visual estimation is popular because it is fast, requires no specialized equipment, and can be adapted to plants of various growth forms. A disadvantage of visual estimation is its subjectivity, making it hard to maintain consistent and accurate measurements.

Estimating cover can be demanding work. Not only do you have to estimate plant cover, you have to be able to identify plants to species (or to whatever category you are using) and find all cover of each species within the quadrat. You will learn not only the process of visually estimating cover, but some techniques for overcoming the drawbacks of subjective estimation. These tricks of the trade make your work easier and more reliable.

The leaves of different species often overlap. This means that the sum of the individual species cover values can exceed 100%. This also means that you have to look beneath other species when recording cover.

A short glossary of cover terms

You will often hear the term "total cover." Unfortunately, just as often it is unclear what it means! In this course I have tried to use three separate terms for different aspects of "cover."

  • Total cover is the cover of all plants, ignoring what leaf belongs to which species. Total cover can take values of 0% to 100%.
  • Combined cover is the sum of the cover values for different plant species or groups. For example, you might measure the cover of each species, then add them together to get combined cover. In one of the course projects, you will measure the cover of different plant groups, and later add them together to get combined cover. Because different species or groups can have leaves that overlap, combined cover can take values that exceed 100%.
  • Overall cover is cover across the entire study area, usually estimated by the average of individual cover values taken from quadrats, lines, etc. For example, total cover in three quadrats might be 60%, 100%, and 50%, leading you to estimate overall total cover as 70%.

Zone of influence

Vegetation scientists recognize that cover in the strict sense is virtually impossible to estimate visually. Leaves and stems of a plant often do not completely obscure the ground. Although these small gaps between leaves technically should not count towards estimates of cover, their large number and small size makes it impractical to account for. Instead, most vegetation scientists apply the "zone of influence" approach. (Warning: Most students, when they first hear of this approach, think it is pretty flaky. But take my word for it—with the proper care, the approach can work very well.)

What is the zone of influence? It is an imaginary boundary around a plant's crown, filling in minute gaps within the crown and smoothing its boundary. This concept is best understood with a diagram. Estimating the cover of the zone of influence is much easier than estimating the cover trying to account for each leaf, stem, and gap. The trick, of course, is being able to define a zone of influence that is both meaningful and consistent. Zones should fill in small gaps in the crown, but exclude large and important gaps. What makes a gap important depends on the vegetation being studied and the study objectives. Similarly, how much the zone of influence smooths the perimeter of a crown depends on how important irregularities of shape are. In practice, vegetation scientists have their own versions of how to define zones of influence, developed on their own or learned from their professors.

Here are two examples of how I apply the zone of influence approach (the red lines show the zones of influence). Notice that the plant at top is relatively simple in outline, with a few simple leaves clustered together. As a result, the zone and the plant border are nearly identical. The larger plant in the middle (Emmenanthe penduliflora) is an entirely different story. The many nooks and crannies around the plant's strongly dissected leaves and fan-shape rosette would drive any cover-estimator crazy. Besides, the plant really does have a zone of influence beyond the limits to its leaves. If I were to estimate cover for this plant, I would estimate the cover within the red line. Wouldn't you?

Plants With zone

(Photograph courtesy of Brother Alfred Brousseau, St. Mary's College)

What the zone of influence is not: The technique of using the zone of influence is not a short-cut method for estimating the true cover of a plant's foliage.  Foliar cover and the zone of influence are two separate ways of measuring a plant's abundance.  The cover of the zone of influence will always be larger (or equal to) a plant's true foliar cover. 

No matter how you develop your own system of zones of influence, it is imperative that you apply the system consistently throughout a study. You will hear more on this shortly.

Building up vs. dividing down

Dividing

"Dividing down" a quadrat

Sometimes it is easiest to estimate cover by thinking directly what proportion of the quadrat is being covered. This works well when cover is consolidated (not scattered over the quadrat) and cover is between 15% and 85%. The figure shows how this system works. In your mind you divide the quadrat into halves or quarters (or sometimes eighths). In this example, you can see that the plant almost covers one quarter, and the cover that overlaps the edges of the quarter is about the size of the uncovered area within the quarter. Therefore, you can estimate the plant's cover to be around 25%. (Remember, you ignore the parts of the plant outside the quadrat.)

The dividing technique does not work well when plants are scattered across the quadrat. In this case, it is more effective to estimate the cover of each patch or plant, and add them together for the complete cover of the species in the quadrat. In the figure, my estimate of the species's cover is 0.7% + 0.7% + 0.7% + 0.7% + 0.7% + 0.7% + 0.4% = 4.6%.

"Building up" cover within a quadrat

Adding
Minus

Sometimes a species will cover almost all of a quadrat. In this case, it is often easier to estimate the cover of the parts of the quadrats without the species, and deduct that value from 100%. In the figure, the species fails to cover about 6% of the quadrat, so the species's cover is about 94%.

As you can tell from these examples, measure cover within the quadrat, even if the plant is rooted outside the quadrat. Similarly, do not measure cover that is outside the quadrat, even if the plant is rooted within the quadrat.

An example

Consider a cartoon plant in a sqaure quadrat frame.  One technique for measuring its cover is to draw lines in the quadrat that divide it into quarters and eighths. Sometimes it is best to draw vertical and horizontal lines.  In this example, it works better to draw diagonal lines. You can do this in your mind or with temporary markers like surveyor's pins or flag stakes.

At this point, you can estimate how much of each eighth is covered by the plant. Alternatively, you can mentally fit the pieces of the plant below the diagonal into the empty spaces above the diagonal.

Many other ways of estimating cover are possible.  Find the method that works best for you.

CartoonPlant

 

 

Advice and practical guidelines for visually estimating cover

Be systematic. Plants within a quadrat will vary in size, shape, and color, making it hard to search for all of them simultaneously. One good approach is to start with making a species list of the quadrat. Just look throughout the quadrat, noting species on your data sheet as you see them. This searching step familiarizes you with the quadrat and using the species list prevents overlooking a species when it is time to estimate cover.

Use calibration templates. A problem with the subjectivity of visually estimating cover is that estimates can vary systematically between investigators. Estimates also vary over the course of a day of field work as folks get tired. A good way to keep cover estimates calibrated is to use templates of known size. You can construct sturdy templates from thick paperboard, adding plastic lamination to make them sturdy enough for field work. The most important part is to make sure that you cut the templates to exact measurements. A range of templates that represent 1%, 5%, and 10% cover work well in a variety of vegetation types. (Of course, the size of a 1% template depends on the size of the quadrat being used!) We keep our templates in our hands when measuring cover so we can calibrate ourselves continually through a field study. Templates

 

One clever student made negative-space templates.  These templates both show standard cover amounts (0.5% on the left, 1% on the right) and hide the surrounding plants, making it easier to focus on estimating the cover of your plant of interest.  Now why didn't I think of that?

(Photos courtesy of Wendy Peterman)

 

TemplNeg1TemplNeg2

TemplNeg3

Don't force yourself to round. Because estimating cover is subjective, even imprecise, many students (and even senior scientists!) feel compelled to record their estimates in "round numbers." They feel that estimating cover to be something like 32% implies that you can tell the difference between 31%, 32%, and 33% cover, and it is true that no one can estimate cover this precisely. But consider what is gained and lost by changing your estimate of 32% to a round 30%. No one but project staff will see the raw field data, so there is no danger in implying a measurement precision not present in the data. That is, nothing is gained by rounding. What is lost is accuracy. If 32% is your best guess, forcing yourself to round to 30% adds an error of 2%! My recommendation is to use your best estimate, period. If you feel better about an estimate of 19% cover than an estimate or 20%, write down 19%. If you feel better about an estimate of 3.7% cover than an estimate of 4% (or 5%, or 0%, depending on how far you round), write down 3.7%.

Don't use cover classes. Cover classes are a favorite device of many vegetation scientists and vegetation science textbooks. Cover class systems divide possible cover values into a few categories, like 0%-10%, 10%-30%, 30%-60%, and 60%-100%. Cover class systems are supposed to save time in the field. After all, it is faster and easier to put an abundant plant in the 60%-100% class than to decide if its cover is 82% or 86%. But cover classes have several very bad characteristics. First, classes cannot be averaged, so you cannot calculate means and standard deviations across your quadrats. That means you cannot apply standard statistical analysis. Users of the system try to solve this huge problem by replacing the cover class with an arbitrary value, like the midpoint of the class. This causes the second problem, because the midpoint of the class is likely to be unrepresentative of the actual cover. For example, there are many more instances of very low cover (0%-1%) than there are instances of cover from 9%-10%. When they get lumped into the same 0%-10% cover class, and represented by the midpoint of 5%, you arbitrarily add a positive bias to your estimates. Who needs that?! This problem is closely related to the rounding problem discussed in the previous paragraph. It makes better sense to write down what you think the cover really is, instead of forcing your estimate into an arbitrary cover class. The third problem is that slight errors at the margins of cover classes can lead to huge differences. For example, if you estimate the cover to be 61% when it is really 59%, it is no big deal if you are using the straight numbers. But if you are using the cover class system, choosing between the 30%-60% cover class and the 60%-100% cover class is a big deal. Finally, in my experience, persons properly trained in estimating cover (a person like you!) can work almost as fast as someone using cover classes. Cover classes are a bad invention-don't use them.

Work as a team. When working in dense vegetation, it is easy to overlook plants (or entire species!) and get poor estimates of cover. In these cases, working with a partner is good idea. A partner can look at a quadrat from a different angle, seeing plants and their cover that might be overlooked from a single vantage point. A partner can also be a data recorder, so the other person doesn't have to change focus constantly between the quadrat and the data sheet. Team members should come up with their estimates privately, so their measurements are independent. If the separate estimates are close, you can feel confident that a consensus value will be close to the true cover. If the estimates are far apart, the team needs to reexamine the plot to see where the problem arose. Although working individually can be faster, the higher quality of results makes working as a team better.

 

Working in teams has several advantages. Looking at the quadrat from two viewpoints makes it harder to overlook plants. Coming to a consensus for cover values improves accuracy. In the photograph, one crew member is pointing out an overlooked occurrence of a species. Notice the template in the hand of the second crew member.

Measurement team

Cross-calibrate. When more than one person is estimating cover, it is important to cross-calibrate. You can do this by starting each day looking at a practice quadrat and estimating cover individually until your numbers agree. If you're working in the same general area over several days, you can set up a reference quadrat. Go back to the reference quadrat at the start of each day, sometimes several times each day, to make sure your estimates of cover don't start sliding higher or lower over time. Cross-calibration is important for two folks working as a team, as well as for several teams working independently.

Sometimes the project leader has more experience estimating cover than the others. In this case, it makes sense for new members to calibrate themselves against the more experienced person.

Putting your knowledge to the test

This is a good time to complete the exercise Visual estimation of cover. That way I can grade and return the exercise with comments by the time you are ready to tackle the bigger exercise (Measuring cover in communities) and the field projects (Measuring cover in the field and Measuring stand basal area).

Making and using quadrats

Size and shape

There are no "rules" dictating quadrat size and shape. But using the wrong size or shape will drive you to distraction! The key is to pick a quadrat size and shape that is efficient to use in the field, producing the most information with moderate effort.

Quadrat size

There is no perfect way to pick the best quadrat size. Some texts suggest quadrat sizes for different types of vegetation. But it is better if you understand the issues behind these recommendations, so you can make the right decisions about quadrat size for your own studies.

Quadrat sizes suggested by Cain and Castro (1959) for sampling different life forms (after Barbour et al. 1998)
Life form
Mosses
Low herbs
Tall herbs or low shrubs
Tall shrubs
Trees
Quadrat area (m2)
0.01-0.1
1-2
4
16
100

It most studies, there is a tradeoff between size and number of quadrats. That is, the more time you spend at each quadrat, the fewer quadrats you will be able to sample from. Decisions about quadrat size depend on this tradeoff, on the need for good interspersion, and on practical concerns in the field.

Consider two types of vegetation studies. In the first, imagine you are studying grass seedlings in Willamette Valley prairies, with many of the plants so small that you will need to put your eyes very close to the ground A 10-m2 quadrat would be impractical in such a study because it would take so long to sample each one that you would have time for only a few quadrats. With just a few quadrats, it is likely that your samples would have poor interspersion. Another difficulty is that it would be nearly impossible to keep track, while examining 10 m2, of what parts you had already searched. A smaller quadrat size would be much better in this case, something like 30 cm by 30 cm shown at right.
Small quadrat

 

Now imagine you are studying the conifer woodlands of southwest Oregon, with scattered trees over a dense shrub layer. How well would a 30-cm by 30-cm quadrat work in this vegetation? The plants are so big—and the walking so difficult—that small quadrats would be highly inefficient.
Shrubs

These examples illustrate several guidelines for selecting quadrat size. If movement from quadrat to quadrat is quick and easy, use a small quadrat. This maximizes the number of quadrats you can sample, which is good for interspersion and replication. If movement is difficult, it makes sense to stay in one spot longer and measure within larger quadrats, as long as the quadrats aren't too large for the size of plants you are measuring.

There is another consideration when you expect to take measurements in the future from the same quadrat. This happens in monitoring projects, or in experiments with measurements before and after treatments. In these cases, you should minimize the disturbance to quadrats; after all, you don't want the process of sampling to affect your results. In dense vegetation, where trampling is a problem, quadrats should be small enough for you to be able to stay outside them while taking measurements. For vegetation dominated by small herbs, 0.5 m is about as far as one can lean over a plot. For these reasons, we use 0.5-m by 1.0-m quadrats in our experimental studies of prairies.

Nested quadrats

Matching quadrat size to plant size makes sense if all the plants are the same size. But many vegetation studies look at plants of many sizes, from herbs and bryophytes to trees. A solution to the problem of which quadrat size to select is to use a series of nested quadrats of different sizes. For example, for any sampling point, you could use a 30-cm by 30-cm quadrat for measuring small herbs and bryophytes, a 1-m by 1-m quadrat for measuring large herbs and tree seedlings, a 10-m by 10-m quadrat for measuring shrubs, and a still larger quadrat for measuring large trees. The added complexity of nested measurements is outweighed by the greater efficiency of using the right quadrat size for the right plants.

Quadrat shape

Quadrats are typically square, rectangular, or circular. The choice is more for practical reasons than scientific.

Circular frames are nearly impossible to construct yourself. You might be able to get a machine shop to fabricate one to your specification. (For a while in the 1950s, hula hoops were popular with vegetation scientists as circular quadrat frames.) Although circular quadrats work well for measuring density, frequency, and biomass, it is difficult to estimate cover inside a frame with curved sides.

You can also define a circular quadrat with a radius cord. Measure a cord to correspond to the area you want to measure. For example, a 25-m2 circular quadrat has a radius of 2.82 m. Running the cord from the fixed quadrat center defines the perimeter of the quadrat. This technique works especially well for trees.

The photographs show how a radius cord works for defining a circular quadrat. (In this case they are using a meter tape, which is slight less convenient.) For measuring tree DBH, the issue is only whether the tree is "in" and should be measured (the tree center is within the quadrat) or "out" and should be skipped (the tree center is outside the quadrat). For most trees, it is easy to see in and out. You need use the cord only for trees along the perimeter.

This tree is
inside the quadrat.
Inside

This tree is outside.

Outside

 

Small frames for square or rectangular quadrats are easy to make. Their ease of construction and use makes these shapes the most popular. I prefer rectangular quadrats, simply because it can be more difficult to see all the way across a square quadrat than across the short dimension of a rectangular quadrat of the same area.

Frames don't work well for larger quadrats. In these cases, rudimentary surveying is needed to mark the quadrat boundary. The surveying technique is described in a later section.

How to make a rectangular frame

Quadrats are usually marked in the field with a piece of equipment called the quadrat frame. There are many versions of quadrat frames, some of which are commercially available. But one of the best types of quadrat frames is one you can make yourself out of PVC pipe. PVC quadrat frames are inexpensive to make, can be constructed in any size, and are lightweight and easy to pack into your study area.
Frames

These instructions tell you how to make a 0.5-m 1.0-m quadrat. After you read the instructions, make a frame of your own to use in your field projects later in the course.

You can get the supplies and tools you need from many large hardware or home supply stores. You will need a little over 3 m (10') of PVC pipe. Half-inch schedule 40 pipe works well, but other sizes and types of PVC will also do the job. You will also need four elbow joints for the type of pipe you purchase, and a small quantity of PVC cement (optional). You will also need something to cut the pipe. Almost any saw will work, but a rigid, small-toothed saw (like a hack saw) works particularly well. There are even large snips made to cut PVC pipe. Finally, you will need a measuring tape or yardstick.

  Cut the PVC pipe into two lengths of 1.0 m and two lengths of 0.5 m. Be careful in your measurements and cutting, because with luck you'll be using this frame in the field for years to come. Pipe and meter stick

Use the four elbows to assemble the four lengths into a rectangle. Make sure the pipe is pushed all the way into the elbows. Adjust the connections so the frame lies flat. Because of the way the elbows are constructed, the inside dimensions of the resulting frame are exactly the lengths of the connected pipe. Measure the frame to confirm this.

You can now use the frame. Just take it apart to carry it into your study area and assemble it on site. But watch out-- PVC elbows have a way of disappearing. It is best to cement them onto one pipe each. Do not cement an elbow to both of the pipes it connects, because then you can't disassemble the frame. Before cementing, double-check that the assembled frame lies flat. Then cement two elbows to each of the short lengths.

 
The unassembled version is easy to pack into the field.
Four pieces

How to place a rectangular quadrat frame in the field

There are some surprising subtleties in placing a quadrat frame in the field. Simply dropping the frame from above is likely to crush vegetation, leading to incorrect measurements and harming the plants you are trying to measure. A better way is to assemble the frame in the vegetation.

First, look at the wrong way to place a quadrat. Note how plants get crushed. Click here for the exciting video, "The Wrong Way."

Now, look at the right way. Crew members insert the sides of the frame through the vegetation at ground level, then connect the corners in place. Done right, no plants get crushed. Click here for the exciting sequel, "The Wrong Way 2: Doing it Right."

This technique doesn't work for welded metal frames, for example, another reason to use the collapsible PVC quadrat frame I recommend.

Special note for 2009

OSU lost the two video mentioned at left!  The links go to an error page.  Sorry!  You can get a sense of what the videos demonstrated from the descriptions.  Here is some more narration:

The first video showed me dropping a constructed quadrat frame onto some prairie vegetation, crushing the plants and making it hard to see what should be in the quadrat and what should be out.

The second video showed a friend and me carefully but quickly assembling the frame through the same vegetation.  It was beautiful!  When you make your own PVC quadrat frame, you'll see how well it works to put the sides through the vegetation.

How to mark out larger rectangular quadrats

Frames don't work well with quadrats of large size. In these cases, rudimentary surveying is needed to mark the quadrat boundary. First mark the initial quadrat corner with a stake. Then run a meter tape from the plot corner. (You'll learn the proper way to locate quadrats in the field later in the class.) At the distance corresponding to the first side of the quadrat, place a second stake to mark the second corner. Then sight a 90-degree corner with a compass and run the tape out in that direction for a distance corresponding to the quadrat's second side. Mark this as the third quadrat corner. Continue like this until you get to the fourth corner, which should be (if you're lucky and/or good) where you started from.

The process is really much simpler than it sounds. Click here to see a cartoon of this process. Just close the window when you're ready to return here.

Even using a good compass carefully, after measuring four sides and sighting four corners, it is hard to end up where you began. A more precise (and more time consuming) technique is to triangulate. That is, you find the next corner by running meter tapes from two other corners. Start as before by running out the first two sides. Call the side from corner one to two "a" and the side from corner two to three "b". Then check the location of corner three by measuring back to corner one. Thanks to Pythagorus, you can calculate "c," the distance from corner three to corner one: c equal sqaure root of a squared plus b squared. If you don't end up exactly on corner one, adjust the position of corner three.

This process is also easier to see than to read about. Click here to see a cartoon of the triangulation process.

Once you have the third corner fixed, follow the same process for locating the fourth corner, and you're done.

If you have four tapes, consider the Bartlett Variation.  Simply leave in place the tape that measures each side of your area.  Voilá, your boundaries are already marked.  You do have to be super-careful with this variation about running the tapes in a straight line from corner to corner.

==> If you are unfamiliar about or rusty on the technical use of a compass, check out Hints for using a compass. <==

Line and point intercept methods of measuring cover

Quadrats are not the only way to measure vegetation cover. The line intercept and point intercept methods can work very well under certain circumstances.

Line intercept method

The line intercept method works best when measuring the cover of plants with distinct crowns, such as sagebrush in the shrub-steppe. The method works by measuring the proportion of the line being intercepted by the species being measured.

(Photograph courtesy of Charles Webber, California Academy of Sciences)

Photo of shrub-steppe
See the photographs to see this process in action. The objective in this example is to measure shrub cover. First, run a meter tape a predetermined distance, as set in your study plan. (The location of this line should be randomly chosen. You will learn about randomization in later sections of this course. Likewise study plans.) Tape as line
Walking along the transect, record where the shrub crowns intercept the tape. These points of interception are typically called "starts" (when enter into a plant's cover) and "stops" (when you emerge from cover). The photograph shows this process in action.

Recording intercepts

Hand on intercept

The calculations for the line intercept method are better seen in a diagram. Say you are using a 20-m line, and the points of interception with the shrub species you are measuring are from 5.0 m to 9.5 m, and from 17.8 m to 20 m, the end of the tape. (Ignore cover beyond the end of the line.) Diagram for line intercept

In this example, the proportion of the line intercepted is

Equation for line intercept.

Thus, the measurement of cover from this line is 33.5%. Be sure to replicate your measurements by recording cover from multiple lines.

It might seem odd to measure a two-dimensional quantity (cover) with a one-dimensional line, but the line intercept method is statistically unbiased, and can be quite fast in the field. An efficient way to run line intercepts in the field is to anchor your meter tape at the specified location and use a compass to walk the tape in the specified direction for the length of the line. Then record the interception points as you reel the tape back up. Very fast.

The line intercept method also works well with tree crown cover. The principal is the same: record where along the meter tape tree crowns start and stop. The trick is sighting straight up, perpendicular to the tape. Commercially available right-angle levels work best. Alternatively, you can look up through a tube, with a crew member making sure the tube is vertical.

Point intercept method for single species

There are two ways to use the point intercept method for estimating the cover of a single species. In the first, you locate points at random within your sampling universe. Each point is a replicate sample location. You measure cover simply as the proportion of points intercepted by the species being measured. That is, the estimate of cover, p, is p equals interceptions over points In the diagram, the calculation is 6 over 15 equals 0.4 . Thus the measurement of cover from this array of points is 40%, similar to the value from the line-intercept method.
Diagram for point intercept

The data from this version of the point intercept method are binary, either interceptions or misses, and discrete. Therefore a proper calculation of confidence intervals requires a special and complicated approach. I won't burden you with it.  Fortunately for our purposes, when the cover is between 20% and 80% or the sample size is large, the calculations of confidence intervals using a Normal approximation are quite close and much easier to calculate.

The confidence interval for the point-intercept estimate of the cover of a species is

p less t s sub p to p plus t s sub p

You can calculate s, the standard deviation of the data just as you would normally, using 1 for interceptions and 0 for misses. You then divide s by the square-root of n to get sp, the standard error of p.

The key to these calculations with the point-intercept methods is, I think, realizing that it is just like all the other situations. The attribute you are measuring is cover over a point. Because it is a point, cover can be only be 100% or 0%. This is equivalent to recording a 1 for interception and 0 for no interception. You observe this attribute in several observational units, the individual points. To go from these observations to an estimate of overall cover, simply calculate the mean, standard deviation, and confidence intervals over these observations. To analyze combined cover of several species, simply add the estimates for each species within each observational unit to get an estimate of combined cover for each point.

The second way to use the point intercept method is to use clusters of points. A common way to array points is in three rows of three points, because these points can be easily located by using a square frame. The estimate of cover at the location of the array is the proportion of points intercepted. Replication comes in when the array of points is located at random through the sampling universe.

The diagram shows this approach in our familiar hypothetical vegetation, in this case with four replicates of a five-point array. Starting from the left, the estimates of cover from the arrays are 0%, 20%, 20%, and 60%. The average of these values is 25%.
Clusters of points

Point-intercept method for more than one species

Sometimes you want to record the cover of several species at a time. If two or more species have overlapping leaves, there can be two or more interceptions at a point. For example, most applications lower some type of rod vertically down through the vegetation, recording the different species the rod intercepts as it moves down the vegetation profile. (By the way, any part of the plant counts as an interception, not just leaves. So also look for interception of stems, flowers, fruits, bark, etc.)

Avoid the trap that catches many beginning vegetation scientists. Just as an individual species can have no more than 100% cover, an individual species can have no more than one interception at a point. Do you remember the distinction between "total cover", "combined cover", and "overal cover"? Where leaves of different species overlap, combined cover can exceed 100%. Likewise, leaves of more than one species can intercept a point. That means the data for combined cover are not binary (0 or 1). The data for combined cover might contain 0 if no species is at a point, 1 if one species is at a point, 2 if two species are at a point, 3 if three species are at a point, etc. Use these numbers to calculate the mean and standard deviation of combined cover.

What about total cover? Because total cover considers the cover of all plants, ignoring what leaf belongs to which species, total cover can either intercept a point or not. That is, measurements of total cover using the point-intercept method produce binary data (0 or 1). I'm glad that things are now perfectly clear.

A note of warning: It is very important when using the point intercept method to use sharp, not blunt points. Because any device you use to define your points belongs to the physical world, it will not be a true point. What if you used a 1/2" steel rod to define your point? This rod will intercept many more plants than a narrower rod, like one 1/8" across. In fact, any device broader than a single, 0-dimensional point will produce overestimates of cover. Because it is hard to avoid this built-in bias, I reserve the point-intercept method for special circumstances.

Should you use the zone of influence with the line-intercept and point-intercept methods?

The point-intercept method has a special capability. It can measure actual cover (without using a zone of influence) about as quickly as it can measure cover using the zone of influence. If your study objectives dictate that you need actual cover (with no zone involved), then by all means use the point-intercept method where an interception must be with an actual leaf or sttem.

Otherwise, all of the reasons for measuring cover within the "zone of influence" of plants pertain equally well to the line-intercept and point-intercept methods. The line-intercept method can drive you crazy if you don't use the zone of influence.

The application is just as you imagine it. Consider the diagram of plants, with the red line representing a zone of influence around the plants. For the line-intercept method, the "start" is when the line enters the zone and the "stop" is when the line emerges from the zone. Line through vege
For the point-intercept method, the trick is to determine if the point is within the zone of influence. In the example, two of the five points are within the zone of influece. (They are a bit hard to see. Look in the upper left and lower right.) Even though none of the points intercepts an actual leaf, the estimate of cover using the zone of influence technique is 2/5 or 40%. Points in vege

Both of these estimates are consistent with estimates of cover within quadrats, using the zone of influence. In fact, if you use the zone of influence with quadrats, but insist that points hit leaves, the two methods will estimate two different things. (This will be quite relevant in some of the exercises and projects, where you will compare the performances of different ways of estimating cover.)

A final note on measuring cover of overlapping plants

Plants in nature are almost always overlapping. How this affects the measurement of cover depends on whether the overlap is of the same species or of different species. When measuring cover of single species, you look at its overall cover, ignoring overlap. But if a plant of species A is overlapped or obscured by the cover or another species, does that mean that the plant has less cover? The answer is that it doesn't. You count the cover as if the other species were not even there.

What this means in the field is that you have to put nose to plot and move plants to the side to see what lurks beneath.

These same principles apply when measuring cover of plant groups. For example, if you are measuing cover of forbs, graminoids, and shrubs, your estimate of shrub cover is the cover of shrubs ignoring what species they belong to. That is, shrub cover in this case is not the sum of the cover of shrub species, because those shrub species probably overlap. But when measuring the cover of the different groups, like the smaller forbs and graminoids, you must look beneath all the shrubs to see what different plant groups might be hidden from view. This includes checking under brambles and poison oak!

HomeGo to Course InformationNext

© 2007 Mark V. Wilson and Oregon State University