These comments are based on the example experimental design that is described above. Relevance of some comments will vary depending on the experimental design used by your class.
Obtaining snails and maintaining them in captivity: I recommend using snails from the family Physidae because they are well known to detect predators through chemical cues and alter habitat use to avoid predation (Turner 1996, Stewart et al. 1999, Brown 2001, Bernot and Whittinghill 2003, Turner and Montgomery 2003). Physids are generally easy to find; they might be the most abundant and cosmopolitan of all freshwater gastropods, occurring throughout North America, Eurasia and much of Africa (Dillon et al. 2002). Physids are particularly abundant in slow-flowing waters with aquatic vegetation, but occur in almost any freshwater environment (Brown 1997, Brönmark and Vermaat 1998). Diagnostic features of physids include a thin (easily crushed between your fingers), coiled shell with a raised spire, and an aperture (shell opening) that opens to the left when the aperture faces the observer and the spire points toward the sky (see images below; Brown 2001). Physids can be collected from shallow waters by hand, or by examining vegetation collected with a dip net (Brown 2001). Use water-filled buckets to transport snails to the classroom or laboratory. Hold snails in aquaria containing aerated, dechlorinated water. Physids will survive, grow, and reproduce on a diet of goldfish flake food. Juveniles hatch from jellylike, crescent-shaped egg masses that are attached to aquarium walls and other hard surfaces.
Obtaining fish and maintaining them in captivity: See "Generating snail responses to predator cues" (item #3 below) to determine if actual predators are needed for this experiment. If you and your students require or desire predators to produce chemical cues, use molluscivorous fish that consume many snails in a short time period, and therefore produce a strong chemical signal. I use redear or pumpkinseed sunfish (Lepomis microlophus and Lepomis gibbosus, respectively) to generate chemical cues for this experiment (see image below, Page and Burr 1991, and Useful Web Sites to identify these sunfish). Redear and pumpkinseed sunfish readily eat snails and have stimulated habitat shifts among physids in my experiment and other investigations (Turner 1996, McCollum et al. 1998, Stewart et al. 1999, Turner and Montgomery 2003). These fish can be obtained from fish hatcheries (see Useful Web Sites), or from ponds or lakes using a seine or hook and line baited with earthworms. Both species generally survive well in captivity on a diet of snails or commercial fish food. Aquaria containing fish should be aerated continuously, and feces should be removed from aquarium floors every few days using a siphon. Visibly sick fish should be isolated from healthy individuals to reduce disease transmission and mortality rates. Because captured fish generally require several days to resume feeding, fish should be obtained at least one week before running this experiment.
Before obtaining fish, be sure to apply for and obtain legal permission for collecting and possessing vertebrate animals. Generally, use of fish in teaching requires approval from an institutional animal care and use committee. If fish are obtained from the wild, a collector’s permit may also be required by your state’s natural resource agency (e.g., Iowa DNR Scientific Collector's Permit, http://www.iowadnr.com/cs/files/542-1367.pdf).
Generating snail responses to predator cues: Snails in any predator-cue treatment of your experiment must detect and respond to chemical cues for this activity to be successful (see Materials and Methods for descriptions of treatments that can be used in this experiment). Physids exhibit interpopulation variability in behavioral responses to predator cues, and some wild populations might not respond to predation at all (Turner 1996, McCollum et al. 1998, Stewart et al. 1999, McCarthy and Fisher 2000, Bernot and Whittinghill 2003, Turner and Montgomery 2003). However, chemical cues and statistically significant habitat use shifts can almost always be generated from at least one of three sources:
Conduct preliminary trials to identify the signal(s) that elicit a behavioral response in your snails. Often, snails respond to cues produced by crushed conspecifics, and fish will not be required for this experiment (Turner 1996, McCarthy and Fisher 2000). To determine if fish predation can be simulated by manually crushing snails, crush several large snails between your fingers, place them in a large aquarium (at least 38-L capacity) equipped with an air pump, air tubing, and an airstone. A carbon filter should not be present in this or any aquarium used in this experiment because filters eliminate odors and other chemical cues.
Between 10-30 minutes after crushing snails, transfer 2 L of water from the aquarium with crushed snails to a 19-L aquarium containing living snails, dechlorinated water, and structurally-complex habitat (e.g., pile of stones and ceramic tiles; see item #4 – Structurally-complex habitat) that provides hiding places for snails. To avoid overfilling the 19-L aquarium, 2 L of water will have to be removed from this aquarium before the water transfer is made. Just before making the water transfer, and every five minutes thereafter, record numbers of snails visible in underwater habitats (i.e., aquarium walls or upper stone and tile surfaces). These snails are considered vulnerable to fish because they would be easily seen by and accessible to these predators. If snails detect and respond to crushed snails, they will begin to crawl about with increased speed until encountering a habitat they perceive as providing shelter from shell-crushing predators (i.e., undersides of tiles, spaces within piles of stones, aquarium walls above the water line). Consequently, numbers of vulnerable snails will decline dramatically. Increase chemical cue concentration through additional water transfers if snails do not respond within 10 minutes of adding water containing cues from crushed snails. Continue periodic recordings of numbers of vulnerable snails. Don’t conclude that crushed snails fail to elicit a behavioral response until after you have added more crushed snails to the 38-L aquarium, and repeated water transfer procedures several times.
Snails from some populations might not respond to cues produced by dead or injured snails alone. Additionally, even if your snails do respond to crushed snails alone, you may wish to use fish to generate stronger cues, or to enable students to view fish in the lab. To generate a strong fish cue, place several large, healthy fish in a large, aerated aquarium (again, 38 L or larger capacity). Feed fish large quantities of snails 10-30 minutes before conducting preliminary experiments to determine if snails respond to fish-generated cues. Conduct water transfers and record numbers of vulnerable snails as previously described until snail habitat shifts are evident. As a guideline, I observe strong habitat shifts in snails inhabiting 19-L aquaria after adding 2 L of water from a 100-L aquarium that contains 4 redear sunfish (85-105 mm total length) fed 10 physids 10-30 minutes before the water transfer was made.
Finally, verify that the chemical cue concentration you plan to use is sufficient by conducting a preliminary experiment with replication and data analysis (see Materials and Methods for descriptions of example experimental procedures). If snails in a predator-cue treatment do not increase refuge use after exposure to the cue, increase chemical cue concentration in subsequent preliminary experiments until strong, consistent responses by snails occur.
Chemical cues degrade rather quickly, and snails will likely leave refuges a few hours after the last addition of water containing chemical cue is made to their aquarium. Therefore, an actual in-class experiment can be run within a day of completing preliminary trials.
Structurally-complex habitat: Behavioral response to predation in this experiment can be measured through increased refuge use by snails after exposure to chemical cues produced by predators or by other snails that have been injured or killed by a predator. Structurally-complex habitat is the primary refuge used by snails in this experiment, and sufficient quantities of structure must be available for behavioral responses to be quantified. At least 1 L of patio stones or similar objects should be placed in each aquarium of predator-free and predator-cue treatments. This is equivalent to the amount of submerged stones required to displace the water line in a volumetric container from the 1-L to 2-L mark. Snails respond to predation risk by seeking out dark, interstitial environments to hide in. To create this habitat, stack stones in multiple layers. To further improve the refuge quality, place ceramic tiles (15 X 15 X 1 cm) or similar cover on top of this pile of stones. Snails seeking refuge from predation that encounter stones or tiles will generally come to rest deep within this pile of stones, or on undersides of tiles. In either case, these snails would no longer be visible to fish or students.
Setting up the experiment: For best results, this experiment should be set up at least 24 hours before it is scheduled to begin. This is advisable so that snails and fish can become acclimated to conditions in the aquaria and classroom, and sufficient time is available for fish in 38-L aquaria to generate strong chemical cues. Depending on how often your class meets, the instructor might have to complete this pre-experimental setup on their own. See Overview of Data Collection and Analysis Methods for detailed descriptions of example pre-experimental procedures.
Data recording and analysis: For this experiment to be successful, students must distinguish snails that are vulnerable to predation from those that are not, and established criteria must be used consistently by all students recording data. The instructor could use input from student groups to establish criteria for designating a snail as vulnerable or invulnerable. However, since this is so critical to the experiment's success, I often establish vulnerability criteria for students and tell them how to distinguish a "vulnerable" snail in an exposed underwater habitat from an "invulnerable" snail inhabiting a refuge from predation. Briefly, any snail that is visible on aquarium walls (including the aquarium floor), or upper surfaces of stones or tiles is vulnerable to predation by fish. A snail that occurs above the water line, or that cannot be seen by fish because it inhabits the underside of a tile or a crevice within a pile of stones is unlikely to be eaten and is considered invulnerable. I prepare students for data analysis by guiding them through two examples of a paired-sample t test using instructions from Zar (1999) and past results from this experiment (see Overview of Data Collection and Analysis Methods, Student Data Set #1 [*.doc] or [*.pdf], and t-test help box).
Writing the paper: I use a lecture and discussion session and handouts adapted from Morgan and Carter (1999) and McMillan (2001) to help biology majors write their scientific paper (see Appendix 1 [*.doc] or [*.pdf]). Additionally, I demonstrate how to perform an online literature search, and provide students a list of relevant background literature that will help them write the introduction and discussion sections of the paper. Selected background literature are also summarized and discussed in class.
Introducing the Experiment to Your Students: This activity integrates many evolutionary and ecological concepts. I have used this experiment to supplement lecture units on natural selection and conservation biology, in addition to population, community, and ecosystem ecology. Before initiating the experiment, I establish a central theoretical framework by introducing students to the concept of natural selection, as well as related terms including fitness and adaptation. I stress that predator-avoidance characteristics result from evolutionary forces. Ecological consequences of predator-avoidance adaptations, and the importance of structurally-complex habitat, are also discussed at the conclusion of the study. See Questions for Further Thought and Discussion for a list of some specific questions I ask. At a basic level, students will understand that predator-avoidance behavior and structurally-complex habitat benefit individual prey and the prey population by increasing the likelihood that some individuals escape predation until after they produce offspring. However, readings and class discussions designed to promote higher-level thinking also help students gain improved understanding of the importance of predator avoidance to the entire biological community. By enabling prey population persistence, predator-avoidance adaptations facilitate long-term survival of all species that rely on this prey for survival (Begon et al. 1996, Krebs 2001). Similarly, due to complex direct and indirect interactions among species, chains of species extinctions can occur if loss of structurally-complex habitat (i.e., refuges from predation) causes a single prey species to become extinct (Begon et al. 1996, Krebs 2001).
Data Collection Methods Used in the Experiment: It is critical that all students understand criteria used in this experiment to classify a snail as vulnerable to predation. I confront this issue by asking the students to think of themselves as fish that can only eat visible snails that also occur underwater. Therefore, snails located underwater that are visible to the observer at the end of the experiment are recorded as "vulnerable," whereas all remaining individuals are considered "invulnerable" to predation. See Overview of Data Collection and Analysis Methods and item #6 (Data recording and analysis under Challenges to Anticipate and Solve) for criteria used to distinguish vulnerable and invulnerable snails.
Data Analysis: Several different statistical tests could be used to analyze data (see Zar 1999 for examples). Student experience in experimental design and statistics should be considered when making the decision of which test to use. A repeated-measures analysis of variance (ANOVA), with treatment type as the main factor, is very appropriate for analyzing data produced from this experiment. An advantage of this tool is that a single test could be used to analyze all experimental data. Results should yield a significant statistical interaction between time and treatment, due to change in numbers of vulnerable snails in one treatment (i.e., predator-cue treatment) but no change in the other treatment (i.e., predator-free treatment). Although repeated-measures ANOVA could be used to analyze these data in advanced undergraduate and graduate-level classes, I find that most undergraduate students have difficulty understanding the concept of statistical interaction. Additionally, repeated-measures ANOVA calculations are relatively cumbersome, and frequent calculation errors frustrate students. Students conducting this experiment in my general biology course are usually non-science majors. For these students, I use paired-sample t tests to analyze data from this experiment because statistics are easier to calculate, and students appear to have little difficulty interpreting results. To assist with data analysis, I supply detailed instructions and two completed examples of paired-sample t tests (see Overview of Data Collection and Analysis Methods, Student Data Set #1 [*.doc] or [*.pdf], and t-test help box). We work through both paired-sample t test examples as a class before students analyze their own data.
There should be evidence that students were carefully observing snail behavior throughout the experiment. Students should notice that snails not exposed to chemical cues (i.e., the predator-free treatment) behave similarly before and after the experiment begins. In the absence of chemical cues, snails tend to crawl about lazily or remain motionless. Alternatively, if chemical cues are added to aquaria, students should see that snails soon perk up and crawl faster as they initiate a flight response. Snails will come to rest within structurally-complex habitat, or above the water line. Some snails in a predator-cue treatment will wander continuously, apparently never finding a habitat they perceive as safe.
A primary purpose here is to assess a student's ability to interpret data and use results from statistical analysis to objectively make the correct conclusion. Using my examples, students will need to understand how to use a t statistic and a critical t value to determine if results support the experimental hypothesis that predator cues induce habitat shifts in snails.
The objective here is to illustrate that science is a continuous process of generating and testing new hypotheses for purposes of improved understanding. New and important questions emerge from results of previous studies.
A main point here is that prey populations cannot survive unless sufficient numbers of individuals avoid being eaten until producing offspring. If prey cannot avoid predators, the prey will go extinct, and predator populations might also become extinct due to loss of food. There should be evidence that a student has used lecture/discussion notes, ecology textbooks, and/or peer-reviewed literature to formulate a reasonable argument.
An objective here is to stimulate student thought about direct and indirect effects of habitat destruction on community stability and biological diversity. Additionally, I wish to get students to think of a biological community and ecosystem as a functioning unit, and species as individual components of that unit. Using information from lectures, class discussions, and readings, a student should demonstrate their understanding that loss of critical habitat of one species can adversely affect many species through a variety of direct and indirect mechanisms, including reduced food and habitat resources, and altered nutrient and energy flow pathways (Begon et al. 1996, Krebs 2001). Examples of classic and recent primary literature that can help guide discussions on relationships between habitat complexity and abundance changes and biological features of ecosystems include Huffaker (1958), Macarthur and Wilson (1967), Walters (1991), Bart and Forsman (1992), and Haddad (1999). Additionally, Krebs (2001) provides an excellent overview of direct and indirect effects of habitat destruction and fragmentation on populations and biological communities.
One component of each student's grade on this activity is based on his or her effort and level of participation. A student's effort and participation score is based on my overall assessment of an individual's active participation in group discussions, and data collection and analysis. Students are also graded on completeness of data tables and worksheets, and accuracy of calculations, and either a short (approximately five double-spaced pages) paper written in the format of a professional scientific journal (biology majors) or the quality of written answers to questions based on this experiment (nonscience majors).
Most students in my general ecology class are sophomore and junior biology majors. Many of them have had few prior opportunities to use statistical tools to analyze their own data. Additionally, some of these students have not yet written a scientific paper. Therefore, I direct much effort toward helping students feel comfortable with statistics and scientific writing techniques. We complete two examples of a paired-sample t test, and discuss how to interpret results from this test (see Student Data Set #1 [*.doc] or [*.pdf], and t-test help box). Additionally, I provide students with examples of peer-reviewed journal articles, and we discuss what should be included in each section of their own paper. A handout and checklist of scientific paper components assist with this (see Appendix 1 [*.doc] or [*.pdf]). Students are informed that I refer to this checklist of scientific paper components when I grade their papers.
I find that use of experiments and statistics in my non-science majors general biology class stimulates learning, critical thinking, and interest in the subject. I often introduce this exercise by stressing that knowledge of the scientific process among nonscientists is important because it can help them critically evaluate information received from a variety of sources, including marketing agencies, politicians, and the news media. Similarly, for biology majors I guide students through two examples of a paired-sample t test before they analyze their own class data. After the experiment and data analysis are concluded, we address several questions in a class discussion (see Questions for Further Thought and Discussion). We begin with relatively simple questions focusing on the value of antipredator behavior to individual prey, and gradually move toward increasingly complex questions that address the combined importance of predator-avoidance strategies and structurally-complex habitat in promoting persistence of biological diversity and natural resources. Following this discussion, I give students some related questions and ask them to submit written responses to me for a grade.
Some of my Questions for Further Thought and Discussion are specifically designed to evaluate student abilities to transfer and apply knowledge gained from this experiment. In this activity, students learn that snails possess chemosensory and behavioral traits that enable them to reduce their mortality risk. However, through a series of questions and class discussions conducted at the conclusion of this exercise, I look for evidence that students also understand the broader applications of their findings. For example, students should demonstrate an understanding that, in a natural ecosystem, predator-avoidance adaptations of prey promote long-term survival of both prey and predator populations, by ensuring that some prey survive until after they reproduce. Furthermore, persistence of predator and prey populations reduces extinction rates among additional species that directly or indirectly depend on the focal prey or predator species for survival. Finally, students should be able to visualize effects of habitat destruction on biological communities, particularly if lost habitat provided organisms with refuges from predators.
At the conclusion of this activity, instructors can use several techniques to evaluate effectiveness of this exercise in facilitating student learning. Alternative methods are described in Evaluation of Course Reforms: A Primer. In the "minute paper," students are asked to provide a written response to a question that is related to the experiment they have just completed. I prefer to ask a simple question such as: "What did you learn from this activity?" By reviewing responses, I gain an improved understanding of how well students understand basic concepts illustrated through the experiment (e.g., small organisms possess physiological and behavioral traits that enable them to avoid being eaten) as well as their ability to relate results to broader ecological issues (e.g., loss of habitat that provides an essential refuge from predation can result in increased predation rates and eventual extinction of the prey species). The instructor can use the minute paper as an additional student-learning tool by reading and discussing student responses during a subsequent class session. If the instructor finds student responses to be incomplete or otherwise unsatisfying, appropriate modifications can be made to this activity to increase its effectiveness as a learning tool.
This experiment can be translated to larger scales by increasing replication (i.e., using more aquaria and student groups) or by increasing the size of student groups. I have little difficulty using this exercise in a large laboratory class of 28 students. Because snails move slowly, a small class consisting of 5-10 students can also easily collect all required data.
Snails used in this experiment occur almost worldwide in a variety of freshwater ecosystems, so this experiment is applicable to a variety of geographic regions. This is a laboratory experiment, and animals can be maintained and bred in the laboratory. Therefore, this experiment can be conducted at any time of year. Although I have only used physid snails and two species of sunfish in this experiment, it is likely that other snail families exhibit predator-avoidance responses similar to those in physids, and that additional kinds of fish elicit behavioral responses in snails.
This experiment is conducted indoors, and data are collected at fixed stations. Therefore, this activity is ideal for students with physical disabilities.
One of the most effective ways to get young people excited about biology is to allow them to work with live animals and to witness unique behaviors and interactions among different species. In fact, I originally developed this activity for use in a freshman-level non-science majors environmental science course because I wanted to stimulate interest by incorporating live animals into the laboratory curriculum. This activity could easily be adapted for lower grade levels by streamlining or eliminating statistical analysis, or focusing strictly on basic principles of animal behavior, or ecological and evolutionary concepts.