BIOL 208 Lab 6: Effects of Stress on Plants PDF

# Lab 6: Effects of stress on plant growth and reproduction: Harvest, Data Collection and Statistical Analysis ## Overview This lab takes place 4 weeks after Part 1 to allow enough time for the plants to grow and reproduce. Learn how and when to conduct an ANOVA and Tukey HSD statistical test. Har...

# Lab 6: Effects of stress on plant growth and reproduction: Harvest, Data Collection and Statistical Analysis ## Overview This lab takes place 4 weeks after Part 1 to allow enough time for the plants to grow and reproduce. Learn how and when to conduct an ANOVA and Tukey HSD statistical test. Harvest and collect data on your response variables. Analyze and interpret your results. Work on your individual lab report. ## Objectives At the conclusion of this lab, participants will be able to: - Design a manipulative experiment to study the ecology of an individual organism - Evaluate the range of tolerance, optimal performance and phenotypic plasticity of an organism - Analyze and interpret data using ANOVA and Tukey HSD statistical tests - Apply Tukey HSD test results to a graph using letter significance groupings - Compose a formal scientific lab report to communicate your results ## Connections to the lecture material - How do organisms deal with resource constraints? - What is the law of tolerance? - What is acclimation and how does it relate to phenotypic plasticity? - What are trade-offs and why do they occur? ## Experimental Design Refresher - Your experimental design has a control (untreated) group and three groups where your abiotic variable was manipulated to three different treatment levels. You should also have several replicates (multiple individual plants) for each treatment. Your experimental design should also include plans to measure two separate response variables: one variable related to growth, and one variable related to reproduction. - For your statistical analysis you will need to combine your measurements for all replicates of a given treatment, calculate the means and compare the means between multiple groups. You will have four means to compare for each variable studied. ## Statistical Analysis: One-Way (Single Factor) Analysis of Variance (ANOVA) - You already know that a t-test is used to compare two means, but how can we compare means from more than two groups? The answer is by doing a one-way analysis of variance (ANOVA). - Like the t-test, a one-way ANOVA compares the variability among means, but can compare more than two means at once. - Instead of calculating a t-statistic however, the ANOVA calculates an F statistic, which is the ratio of the variation between a group of means relative to the variation within the groups (Figure 6-1). ## Figure 6-1 A. Graphical representation of the types of variation used to calculate an ANOVA F statistic. A) The figure shows means calculated for each group, represented by a straight line drawn through each dataset. The variation between groups is represented by vertical lines drawn from the respective group mean to the line representing the total sample mean (mean of ALL the means). B) The variation within groups is represented by vertical lines showing the distance of each data point to the group mean, represented by a line. The F statistic compares the variation between groups (A) to the variation within groups (B). ## When performing a one-way ANOVA test, the null hypothesis is that - There is no significant difference between any of the groups. The null hypothesis will be rejected if at least one of the means is significantly different. A one-way ANOVA can tell you if you have a significant difference between any one or more of the groups, but the ANOVA does not tell you which groups are different from which. To determine which groups are significantly different, we must also use a "post-hoc" test, which in this case performs pairwise comparisons of the means after the ANOVA test is conducted. In this course, we will use the Tukey Honest Significant Difference post-hoc test. ## For BIOL208 you can use the Excel Data Analysis ToolPak to calculate all the required values for your Single Factor ANOVA analysis: - Click the "Data" tab in Excel - Click on "Data Analysis" Toolpak (see Stats Appendix for how to install the ToolPak). - Select "ANOVA: Single Factor" - Input the range for your complete data set under "Input Range". Make sure your data is arranged in columns. - If you have data labels included in your data range selection, click the checkbox "Labels". - Leave the "New Worksheet Ply" selected, and leave the Alpha as "0.05". - click "OK" - Retrieve the following data from the output table - "F", "Between groups df", "Within groups df", and "P-value". - Format your statistical statement as shown below: - (Fbetween group df, within group df, p-value) - Eg. (F2,12= 10.14, p=0.0026) ## Please see eClass for a video tutorial on using Excel to calculate a single factor ANOVA and the Stats Appendix in this manual. ## Tukey HSD Post-Hoc Test The results from your one-way ANOVA will tell you if there is a significant difference between any of the groups, but does not tell you which ones are different. In order to determine which groups differ, you need to calculate the Tukey HSD Post-Hoc test (Table 6-1). HSD stands for "honestly significance difference". The Tukey HSD test does a pairwise comparison of the means, but does so in a more robust way than doing multiple t-tests over and over. Excel does not have a Tukey HSD Post-Hoc test, so we will use an online stats calculator called Vassarstats (www.vassarstats.net) to run the Tukey HSD Post-Hoc test on our data. ## For BIOL208 you can use Vassarstats.net to calculate your Tukey HSD test: - Go to www.vassarstats.net - Select "ANOVA" → "one-way ANOVA" - Enter the number of means ("samples") you are comparing. Select "Independent Samples". - Enter data set manually and click "Calculate" - Scroll down to the "Tukey HSD Test" Output section. - Interpret the results and mark the statistical grouping with letters above your bar graph. ## Please see eClass for a video tutorial on using Vassarstats to calculate a Tukey HSD test and the Stats Appendix in this manual. ## Statistical Significance Letter Groupings To communicate the results of your Tukey Post-Hoc Test, you will use letters above the means in a bar graph to signify which mean(s) or groups of means are different from each other. ## Means that are not significantly different from each other share - the same letter, whereas a unique letter represents that the means are significantly different from each other. To illustrate how the lettering is determined, look first at the results in your Tukey HSD data (Table 6-1). According to the table, Means 1 and 3 are significantly different from each other, so each data point gets a different letter ("a" for Mean 1 and "b" for Mean 3; Figure 6-2). ## Mean 1 was not significantly different from Mean 2. - So Mean 2 also gets marked with an "a" to indicate it is in the same significance group as Mean 1. Being in the same letter group means there is no significant difference between the two samples. ## Lastly, Mean 2 was significantly different from Mean 3, so those should have different letters. - Which they do, since you already labeled Mean 2 with an "a" and Mean 3 with a "b". By adding these notations to your graph, a reader can know which means are significantly different from each other. In your figure caption it is important to make note that the letters represent significance groupings from the Tukey HSD Post-Hoc Test. ## Part 2: Harvest, Data Collection and Statistical Analysis Procedure *Work on this as bench groups. You will be expected to do an individual lab report on your experimental design and the results you obtain, so take detailed notes! See the assignment information on eClass. 1. Collect your plants for harvesting. They have grown for 4 weeks in a climate controlled growth chamber. 2. First, make qualitative observations of your plants. Do they physically look different from each other? Take extensive notes and take photos to use in your lab report. 3. Measure and collect the data for your chosen response variables: one variable should be related to plant growth and the second variable should be related to plant reproduction. 4. After you have collected your data, clean up by disposing of your plants and soil and wash your pots thoroughly (they must be completely clean and free of contaminants to reuse next term!). 5. Tabulate all the data into Excel. Use the Descriptive Statistics Toolpak to calculate means and 95% confidence intervals for each data set. If you are unsure how to do this, please see eClass for a tutorial, or refer to Appendix II in the lab manual. 6. Graph your data (in two separate graphs, one for the growth variable, and one for the reproduction variable). Use your 95% confidence interval for the error bars. 7. Conduct a one-way ANOVA test using the Excel Data Analysis ANOVA: Single Factor function. You will need to do separate one-way ANOVAs for each variable you measured. (ie. one ANOVA for your growth variable, one ANOVA for your reproduction variable). 8. Record the statistical outputs from the ANOVAS. 9. Use Vassarstats.net to complete the Tukey Post-Hoc Tests for each of your data sets. 10. Use your statistical results to indicate significance letter groupings on both of your graphs. 11. Make use of any remaining time in lab to work on your lab report.