Challenges to Anticipate and Solve
- Low gall density: Gall densities fluctuate from year to year. One to several weeks before the lab is scheduled, check the willow leaves at your site. If gall densities are quite low (students will have to sample many leaves to get any with galls), then adjust sample sizes to include more leaves or look for alternative sites with more galls.
- Early leaf drop: The date of leaf drop varies from year to year. You may need to monitor your study site and adjust the date of the lab so that it can be done before most leaves fall. If you do get caught by early leaf drop, you might be able to adjust the hypotheses tested. If trees are close together, hypothesis 1 canít be tested. Some hypotheses can be tested with leaves on the ground. If your lab occurs partway through leaf drop, you might test the hypothesis that leaves with galls will be abscised earlier than leaves without galls (trees getting rid of herbivores) against the hypothesis that leaves with galls are retained longer (galls are carbohydrate sinks, so trees have more invested in them).
- Choice of hypothesis to test: Students may not be particularly motivated here. Sometimes they just choose the first one, or the one that looks easiest (fewest samples required). A bit of discussion about these hypotheses when you introduce the lab should help motivate students. Encourage students to formulate their own hypothesis based on what they have learned of plant interactions in lecture, the introductory material in the lab handout, or initial observations at the study site. Be sure they have some time in lab to make this decision, and discuss briefly with each team why they chose that hypothesis.
- Statistical literacy: Depending on how much exposure your students have had to statistical testing, they may not be comfortable with data analysis and testing. The examples should help, but you may need to go through some examples on the board after the students collect their data. Ensure them that the calculations are really quite simple, and emphasize the importance of statistical testing (e.g., how big would the differences have to be before you would be confident saying there are differences?). Walk around among the teams and discuss their results with them to be sure they understand whether they should accept or reject their hypothesis, and why.
- Production of graphs for oral reports: Students often have trouble with 2 aspects of producing graphs. The first is deciding upon an appropriate kind of graph for their data. You could discuss this with each group in lab after they have collected their data, or discuss more generally in lecture. You could give examples of the best kinds of graph for each hypothesis. The second issue is actual graph production. Many students are familiar with Excel, and some have used it to graph data before. But they often use Excelís defaults, and donít format graphs appropriately. You might want to give some guidelines about graphing, or post an example on a course webpage (or give a handout), with large enough axis labels, a descriptive figure legend, etc.
- Oral presentations: The quality of oral presentations may depend on whether students have done this before in a biology class. If you have time in lecture, you might want to give them an example. Present an oral presentation yourself, perhaps based on a previous lab students did, so they see what level of detail and professional conduct you expect. Be sure to adhere to the time limits!
Introducing the Lab to Your Students:
I usually try to introduce herbivory as a type of species interaction in the lecture part of the course. I cover the various ways plants defend themselves from herbivores, and how herbivores both respond to and are affected by chemical and physical defenses. During the lab, I give a brief overview of galls as a particular type of herbivory, and show some different types of galls. The Detailed Description of the Experiment provides introductory information.
Activities in the Lab:
- I give a very brief introduction to the 1st hypothesis, which all students will address. I ask students to form teams of 2-3 members.
- For descriptions of aborted galls, successful galls, and parasitized galls, see Clancy et al. (1987). Aborted galls are usually quite smaller than active galls, and tissues are nearly solid (seen when the gall is cut open: Image of Aborted Galls) rather than organized into a hollow structure: Image of Opened Galls). Successful sawfly larvae chew emergence holes (mean diameter 1.3 mm for Pontania sp. in Arizona) that are larger than holes made by parasitoids. To detect parasitized galls, cut open and observe under a dissecting microscope. If the larva has been parasitized, you will typically see an egg or other stage of the parasite (see image of Parasite).
Potential modifications of this lab activity:
- Rather than discussing the class results of Hypothesis 1 in lab, you could have each group include a very brief summary of what the class found, and how this provides relevant background information leading to the specific hypothesis their team tested.
- Students could be asked to hand in the literature citations they used for the oral report.
- Students can submit a formal written lab report (similar in style to manuscripts submitted to Ecology) instead of the oral report. See Written Report Gradesheet [DOC] (56 KB).
- For students with more experience in analyzing data, you might give more general comments on methods (e.g., for hypothesis 2, randomly sample leaves with galls and leaves without galls, recording whether chewing damage is present or not. How many leaves should you sample?), or not give any hints and have students propose methods and types of analyses. You would need more time in lab to inspect their proposals before they collect data.
- This activity can easily be extended to 2 weeks, with more student-based inquiry. During the first week, introduce the students to the study system, have them make their own, unguided observations about galls on willows, and collect data to test hypothesis 1. After summarizing and discussing class data, have each group think about an additional question to ask about the distribution of galls on willow leaves. (Do not provide them with the list of additional hypotheses.) They can propose their own hypothesis based on this question, which you can discuss with them. By the end of the first lab period, teams will hand in a written version of the additional hypothesis they will test, the types of data they will collect, sample sizes, and methods for selecting sample units. The instructor should review these, and provide comments to ensure a viable hypothesis, sufficient sample size, and guidance as to appropriate statistical tests. During the second lab period, teams will collect and analyze data on their hypothesis, and begin preparing their oral reports. An example of a student-generated hypothesis is: Gall densities are higher on willows in close proximity to a water source (e.g., a pond) than on willows farther from surface water.
Questions for Further Thought
Conclusions about hypothesis 1:
- You will probably find that trees do vary considerably in the level of herbivory by sawflies (means differ; but look for large standard deviations; you might mention that statistical testing helps sort out whether these apparent differences in the mean values are real, i.e., statistically significant).
- Potential explanations include that individual trees vary genetically as well as in their ecological conditions, age, etc. These differences could result in differences in nutrient levels, concentrations of chemical defenses, etc. There may also be differences among trees in how many predators or parasitoids are around.
- If willow trees preferentially drop leaves with galls, then trees may not vary much in how many galls are on intact leaves. But you may find higher numbers of galls on leaves on the ground under some trees.
For additional hypotheses tested by individual teams:
- Ecological explanations can be found in the background information supplied with each hypothesis, as well as in the literature cited.
Assessment of Student Learning Outcomes
I find that students often have difficulty with statistical tests. I try to make sure I visit each team several times while they are analyzing their data to make sure they properly summarize their data before beginning analysis, understand the basic idea of what the statistic is testing, and understand how to interpret P-values and test statistic values. Preparation for the oral report should be pretty straight forward if the students read through the Oral Report Gradesheet [DOC] (56 KB) in advance.
Translating the Activity to Other Scales
- Translating this experiment to larger scales: you may want to form larger teams (5-6 students), and assign each team a hypothesis to test rather than taking time to have them select a hypothesis. Larger teams should be able to collect more data, so increase the sample size suggestions.
- Translating this experiment to pre-college settings: you could omit the statistical tests, and revise the hypotheses to simpler terms. Focus on having students understand the basic concept that herbivory is everywhere, and some plants are more susceptible to herbivores than others. Herbivores also have choices in where they place their offspring, and this affects how well their offspring do.