Biology Books |

SAMPLING TECHNIQUESObjectives for Today's Lab. At the conclusion of this lab, (1) you will understand how to use several methods to sample populations of plants and animals including the use of transects, quadrats, and so-called "plotless" techniques. (2) you will learn that one need not measure "everything" to get an estimate of various characters of natural populations. (3) you will have a better understanding of how to devise and implement your own field sampling methods to answer ecological questions about patterns in nature of interest to you. Equipment Needed for Fieldwork This Week. - field clothing. - clipboard, notebook, paper and pencils for recording your data - in addition, you will need the specific equipment on the checklist below in the section on methods. What Your Group Will Hand In At the End of Class. Your group should hand in your data analyses (tables, graphs, etc., detailed below) as well as your original field data. ECOLOGY LAB - ECOLOGICAL SAMPLING: METHODS FOR DETECTING PATTERN IN NATURE -------------------------------------------------------------------------- Introduction. Nature shows pattern. This includes patterns in the distribution and abundance of plant and animal populations along environmental gradients in space, time, or both, as well as community and ecosystem patterns at higher levels. All of these patterns stem ultimately from the ecological and evolutionary pressures on individuals living a complex world. According to Robert MacArthur, a founder of modern ecology, the aim of ecology is to search for, describe, and explain patterns we see in nature. The first step in this process is to describe pattern. Although this might seem to be the easiest step, one of the most challenging tasks in ecology is to devise descriptive techniques that best do this given the limited time and resources available. Since it is impossible to measure everything, the optimal sampling scheme involves judiciously selecting a few indirect measures that reveal the pattern(s) of interest. This is called "SAMPLING", a procedure by which you collect data on a subset of individuals and estimate the relevant components of natural populations, communities and ecosystems to reveal the pattern of interest to you. For example, suppose you were interested in determining the effects of naturally occurring fires on plant community biodiversity. Suppose your interest in this were because you were a range manager in a prairie ecosystem, and you wanted to know what would happen if you were to implement a management policy to put out all fires. One way to find out would be to choose two areas of identical prairie and allow one to burn naturally, and put out every fire on the other. Eventually (probably in a decade or so) one would simply sample the plant communities on each and look for differences that COULD be due to the difference in fire frequency. (In fact, studies have shown that fire is essential to prairie diversity...) The trick, of course, is to come up with a good way to select which individuals to include in your sample that are "truly representative" of the actual number of various species of plants of each kind without having to count every individual. In order to collect representative samples, one must take data on individuals targeted on a "RANDOM" basis. For an ecologist attempting to measure how many grass species are in a prairie, this means that an individual cannot be sampled just because it is conspicuous and easy-to-find (such as a big weed with a big flower). In fact, the biggest species might be among the rarest. Plants must be sampled randomly otherwise the resulting estimate would be a "biased" one. There are almost as many sampling techniques as there are research questions in ecology. However, all of these techniques can be lumped into two main groups depending on how the sampler encounters the small number of individuals on which to make measurements. These are: (1) PLOT OR QUADRAT BASED METHODS which randomly locate a piece of ground within which individuals are sampled, and (2) PLOTLESS METHODS which lead to random encounters with individuals to be sampled along transects without using a marked off grid. These approaches will be discussed in class, research groups will form, and then YOU WILL BE SENT OUT ON YOUR WAY TO GREATNESS. Methods for Today's Lab. Today, following a brief discussion about sampling, you will form small research groups (of 5 people per group), and we will travel to a local park to conduct today's sampling studies. One research team will establish a fairly large plot, or "quadrat," in a representative area and perform a complete census of the trees in their plot. The remaining two research teams will deploy line transects in similar wooded areas and estimate tree species diversity and relative abundance indirectly using a transect sampling scheme called the "Point Quarter System." Pay close attention to the details below of what data are collected in the field and how it should be analyzed and presented. Steps Once On Site: Methods for the group using the "quadrat" sampling scheme. - Establish a starting corner and the approximate area for your sampling quadrat. Ask me for advice on which habitat type in this park you should select (either the "west" or "south" woods at Taylor arboretum). - Use your measuring tape to find the other three corners of your sample quadrat. I suggest that you measure out 30 meters along a side in one direction and try as best as you can to measure out another 30 meter side at a right angle. Continue this on around until you close your quadrat square. Tie little strips of day-glo flagging to sticks or trees at the corners so you can see them. REMEMBER TO PICK ALL FLAGGING UP WHEN YOU ARE DONE. - Proceed as systematically as you can to record every tree in your quadrat. Identify each tree to species (although if stumped you may use oak species 1, oak species 2, etc.), and measure the circumference of the tree at a height of about 1.5 meters. Ignore vines, young thin trees (circumferences less than 30 cm), and dead trees. - If you finish completely censusing your quadrat in less than 45 mins, then mark off another 30 meter square plot in an adjacent area and repeat the censusing. - DATA ANALYSIS - DENSITY ESTIMATION: To estimate density, simply calculate the total number of individuals for each species and divide this number by the total area of your study quadrat. This will give you the density per square meter for comparison with data from the other groups. - DATA ANALYSIS - DIVERSITY ESTIMATION: To estimate diversity, I want you to use several different methods. One is to simply total the number of different species you encountered. Secondly, I want you to refer to Appendix 1 and calculate several commonly used diversity indices from your data. We will compare diversities among these stats and with the diversities from the other groups. Methods for the groups using the "point-quarter" sampling scheme. - Establish a starting point and a direction for your transect, and place a small flag at this point. Ask me for advice on which habitat type in this park you should select (either the "west" or "south" woods at Taylor arboretum). Note: since the starting point is established subjectively, don't sample anything at it. - Extend your measuring tape from the flag you just placed along the direction of your transect and place another flag exactly at the 10 meter mark. This point represents the "origin point" of the point quarter technique. The "quarter" is represented by the four quarters created by dividing an imaginary circle with the origin point at its center into four equal quarters. To help you see these quarters place a stick perpendicular to the transect line at the origin point (I will demonstrate this in the field). Now you are ready to take your measurements at the origin point. - Choose the tree that is closest to the point in each of the four quarters. Write down three things on your data sheet: (a) the species/type of plant (use the field guide and ask if your are unclear), (b) the circumference of the tree at a height of about 1.5 meters, and (c) the point to plant distance. Ignore vines, thin trees (circumferences less than 30 cm), and dead trees. The following illustration may be helpful. quarter C . quarter D . nearest tree o . . . o nearest tree . . . . . . . . . next origin point ... transect line --------------------------O------------------------------>>> ... . . . . . o nearest tree . . nearest tree o . . . quarter B . quarter A - Move ahead another 10 m approximately in a straight line, and repeat the previous steps at least ten more times. - DATA ANALYSIS - DENSITY ESTIMATION: This is a tricky process, so follow the data analysis worksheet closely. - DATA ANALYSIS - DIVERSITY ESTIMATION: same as for quadrat sampling described on the previous page. Some questions for thought for the comparison of point quarter sampling with transect and quadrat sampling exercise. - How did the results of these two sampling methods compare? - How did the "collecting effort" differ among these two? - Which method would be better for what type(s) of ecological questions? - What might account for any observed differences among habitat types in the data collected among all groups? - What might account for observed variation among sampling stations along the same line transect? IT IS ESSENTIAL TO THE SMOOTH RUNNING OF THIS AND SUBSEQUENT LABS THAT YOU MAKE SURE YOUR EQUIPMENT IS RETURNED, CLEANED, AND NEATLY ORGANIZED SO THAT THINGS ARE USABLE FOR THE NEXT LAB. HAVE FUN!!! Data Sheet for Quadrant Sampling. Names of research team _____________________ Quadrat total area _______________ Species Circumference (cm) Species Circumference (cm) ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ ------------------------------ Data Analyses for Quadrat Sampling. Names of research team _____________________ Quadrat total area _______________ Species Total Number Density Proportion of All on Your Plot trees/m} Observations, pi -------------------------------------------------------------------- all species DDD> 1.00 -------------------------------------------------------------------- -------------------------------------------------------------------- -------------------------------------------------------------------- -------------------------------------------------------------------- -------------------------------------------------------------------- -------------------------------------------------------------------- -------------------------------------------------------------------- -------------------------------------------------------------------- -------------------------------------------------------------------- -------------------------------------------------------------------- -------------------------------------------------------------------- Notes: List species in descending order of occurrence. The column "Proportion of All Observations, pi" is the fraction of all trees you censused that are of each species. E.g. if you censused 80 trees in total and 10 were holly, then pi= 10/80 = 0.125 for holly. You will need this column for your diversity calculations. -------------------------------------------------------------------- Diversity Estimations (see appendix 1 for calculation steps) Total Number of Species _________________ Shannon Diversity Index, H' __________________, exp(H') __________________ Evenness of Shannon ________________________ Simpson Diversity Index, Hs __________________ Data Sheet for Point-Quarter Sampling Scheme. Names of research team _____________________ Point # Species Circum. Distance Point # Species Circum. Distance (cm) from Point (cm) from Point ------------------------------------ ------------------------------ (1) A (6) A B B C C D D ------------------------------------ ------------------------------ (2) A (7) A B B C C D D ------------------------------------ ------------------------------ (3) A (8) A B B C C D D ------------------------------------ ------------------------------ (4) A (9) A B B C C D D ------------------------------------ ------------------------------ (5) A (10) A B B C C D D ------------------------------------ ------------------------------ Data Analysis for Point-Quarter Sampling Scheme. Names of research team _____________________ Total number of point to tree distance measurements ____________ = N Sum of all point to tree distance measurements ____________ (m) = d D _ Average point to tree distance (d D / N ) ____________ (m) = D _ _ _ Average area occupied by an average tree ( D * D ) ____________ (m) = D} _ Total density of trees in the habitat (1/D}) ______________ trees/m} Now, fill in the table below for each species: Species Total Number Density Proportion of All Among all Points trees/m} Observations, pi -------------------------------------------------------------------- all species DDD> 1.00 -------------------------------------------------------------------- -------------------------------------------------------------------- -------------------------------------------------------------------- -------------------------------------------------------------------- -------------------------------------------------------------------- -------------------------------------------------------------------- -------------------------------------------------------------------- -------------------------------------------------------------------- -------------------------------------------------------------------- -------------------------------------------------------------------- Notes: The basic idea (as described by Cottam and Curtis, 1956) is to estimate the area occupied by an average tree, expressed as square meters per tree, and noticing that the inverse of this quantity has units of trees per square meter. One finds the average area occupied by trees by assuming that the average point to tree distance is related to the average tree canopy diameter. In fact, the relation is assumed to be the exact square of the average point to tree distance (mean area = D} ). To find the tree density (number of trees per m}) for each species for the point-quaters method, simply multiply the total density of trees that you found among all trees by the pi values for each tree species. ------------------------------------------------------------------------- Diversity Estimations (please use the worksheet above for quadrat sampling because there's no more room on this page. Appendix 1 - Calculation of Diversity Indices