Thread: Quadrats for ecological sampling

What is the technical definition of a quadrat? Some sources I've read imply that a quadrat is a single square frame. Other written sources describe a quadrat as a square frame that is divided into equally sized subdivisions. Does the word 'quadrat' alone by definition perhaps refer to a single square frame that is considered to be a basic quadrat (or classic, if you like)? Is the frame that is divided into subdivisions (I've also seen referred to as a 'frame quadrat') therefore considered to be a type of specialized quadrat, perhaps utilized for extra precision in the practice of recording the abundance of a species being sampled? I am aware that there are also 'point' quadrats, but these I know are very different in model to the aforementioned quadrat frames (single or subdivided) and therefore something of their own (although used for the same purpose).
I really don't know which type of frame to picture when I read about quadrats. I wonder whether the word quadrat (that is without another preceding word like 'frame') refers exclusively to the single frame version, or if the word is used in a relaxed way and could mean either single or subdivided (aka. frame).

My confusion also brings me to something in a book I read about estimating population density. The book states:

Estimating population density - a method for plants and small mobile animals:

Place quadrats randomly

Count the number of organisms of the species you are investigating in each sample quadrat

Find the mean number per quadrat

Estimate the population size using the formula:

mean number of organisms x area of field/area per quadrat

The part of this that puzzles me is "find the mean number per quadrat." Because I don't know what type of quadrat to visualize, I'm unsure what you would do to obtain the mean. Would the quadrat in question not have to be one divided into subdivisions so that you would therefore carry out the equation: number of organisms in all subdivisions/number of subdivisions ?

Although this seems most sensible, I'm confused because the preceding text in this particular book seems to be discussing quadrats of a single frame only, and refers to them simply as 'quadrats.' In fact it isn't until the subsequent page of the book that 'frame quadrats' (i.e. quadrats that are subdivided) are introduced to the reader and discussed for the first time.

How then would you "find the mean number per quadrat?" Would you carry out repeats? In the case of plant species sampling, wouldn't you just keep getting more or less, if not exactly, the same data figure?

I'd be very appreciative if anyone might kindly shed some light on this!

The word 'quadrat' does, indeed refer to a simple square frame. It need not even be physical. The quadrat might be a shape on the ground measured with a tape measure or similar. When you use physical quadrats, they might be subdivided into smaller squares. The quadrat in that case depends on which square you use for the count. If you count all organisms in the larger square, that is the quadrat. If you count the organisms in a smaller square only, the smaller square is the quadrat.

The size of the quadrat you use will vary according to the species you are counting. If you were counting deer from a helicopter, you might use an imaginary quadrat of an estimated one kilometre square. If you were counting microscopic rotifers, you might use a microscopic grid seen through the microscope.

You will use many quadrats, placed at random in the area you are researching. The helicopter deer count might use (for example) 25 imaginary quadrats of one kilometre square over an area of 1000 square kilometres. You would use a map and random numbers to decide on 25 places for your imaginary quadrats, with the random numbers determining map coordinates.

On the other hand, I have seen researchers with small quadrats counting soil animals decide where to put the quadrat by throwing a dart over their shoulder with eyes closed.

In theory, a quadrat does not even have to be square. It is square by tradition more than necessity.

Thank you very much, Skeptic, for such a informative and enlightening reply! I had no idea to the degree of application the quadrat model has in ecological sampling. The literature I've read only discuss the use of physical quadrats in the sampling of plants and small animals with minimal locomotion. It's truly fascinating that the same sampling method and principle applies in both microscopic studies and the studying of much larger animals like deer, involving such great scale. Your example of sampling from a helicopter by this method is very revealing. I'd be keen to learn more about this. I have indeed previously wondered about the means by which large species groups are surveyed from the sky when seeing brief footage of this being carried out in documentaries.

In regard to the subdivided quadrat: is it more commonplace to count the organisms present in every square subdivision of the grid, or to select just one (such as the smaller, or larger square that you refer to) or a few of the squares from the grid to record data from? Are there quadrats subdivided into equally sized squares, and also ones that are subdivided into squares of different sizes? I'm aware of the two common methods of estimating species abundance. Both involve using all of the squares of a subdivided frame quadrat consisting of equally sized squares.
The first method is to estimate the percentage cover of a species by counting the number of equally sized squares the grid consists of and then working out the percentage of the overall grid that each square constitutes. You would then be able to estimate the percentage cover of the species by counting the number of squares of the grid the species is found in and multiplying that number by the percentage each equally sized square constitutes of the overall frame.
The second way of estimating abundance, that I frequently read about, is to measure the frequency of the species by counting the number of squares of the frame the species is found in, dividing this by the number of squares the frame grid consists of altogether, and then multiplying the answer by 100 to obtain a percentage.
Because these methods are both common practice, is it therefore most usual to record data from all of the squares?
Which type of quadrat might you guess the author of the book is referring to in the extract "estimating population density" I posted? I'm very confused as to how that mean can be obtained!

I'm amused and again fascinated that researchers select random locations for sampling by throwing darts over their shoulder! Am I correct to understand that quadrats are always used in random sampling, but transects are used systematically in applications where you wish to study frequency, abundance and distribution of organisms of a species? You may thereby find correlations between environmental factors and the distribution of certain organisms.

I'm very grateful to you for sharing your knowledge on this subject, and so helping my understanding.

j.r.

One of the advantages of quadrat techniques is their flexibility. It would be wrong for me to try to lay down strict rules for you, since real field researchers vary their methods so much. Ultimately, how you do your quadrat count has to be using a method that can be related mathematically to the whole area you are studying.

If you have a large quadrat - say a metre square, that is subdivided into smaller squares, the exact squares you use will depend on what you are counting. If you are counting, for example, sea snails, you may decide to use the whole square. If you are counting, however, ants, you might need to use a small square, since there are too many ants to count in such a large area. You would not record data from all the squares in this case, but from a sampling of at least 20 such counts, with each small square placed according to a randomising system. Each count comes from the quadrat - large or small - being re-set in a position determined by random numbers or other effective method of obtaining randomness.

As I said, the quadrat does not even need to be square, if the count can be multiplied out to a number representing the entire population of that ecosystem. It can even be imaginary, like lines drawn on a map. Some counts are done using satellites (like counting whales) on an imaginary grid. Still the same method. You can vary the method many ways, as long as you use excellent mathematical good sense.

On the extract "estimating population density", it depends on what species you are measuring. Snails, ants, whales or people?

Whether you use quadrats or transects is more a matter of practical necessity. I have, for example, used transects when making counts underwater in a strong current. It is easier to anchor a transect line than to carefully position quadrats randomely, when a strong current stuffs everything up!