Classical scientific methods include hypothesis forming and testing to resolve scientific questions and build theories. A hypothesis might be tested repeatedly and thus evolve into a theory. Generating and testing hypotheses provides important practice in risk taking and divergent thinking. Ecological experiments are used frequently by ecologists to answer both theoretical and applied questions (for example, see the article Designing Environmental Field Studies by Eberhardt and Thomas).
Experiments in ecology follow the logical pattern of deductive reasoning used in other scientific fields. Although many science teachers have their students collect data without using experimental techniques (called "mensurative" experiments), the majority of ecologists consider that only manipulative experiments, when you assign treatments and controls and where replication of each treatment plot are used, to be preferable. An experimental approach is better for producing credible, reliable scientific data that can withstand peer review and guide natural resource management.
There are 5 key features of science research as a mode of inquiry (AAAS, 1989):
- The process of forming and testing hypotheses
- Developing an experimental design
- Obtaining evidence by observation and measurement
- Using logic and insight to analyze data
- Developing an explanation based on valid observations using logic and application of conceptual knowledge
1- Process of forming and testing a hypothesis. Why do you need a hypothesis? Questions are the basic of inquiry; without a hypothesis, you don’t have a framework to test your data. A research question asks: "What relation exists between two or more variables?" It expresses a possible, clearly stated relation between these variables, and the question implies possibilities for an experiment to test the relationship. Research questions are often stated as predictions of a relation between two or more variables. Predictions are synonymous with hypotheses since you must have some sort of expectation of an outcome or result when you form a hypothesis. This also implies that you already know something about the ecosystem, the particular site and the existing conditions there.
There are a wide range of questions that you might investigate as ecologists. Overall questions include: What factors determine the abundance of species? How important are particular interactions between species? Question investigated in plant ecology, for example, focuses on the patterns, causes, and consequences of plant abundance and/or distribution in nature. Once you have your research topic and overarching purpose, such as an investigation of nearby forest, clearly in mind, you can start framing your research questions and generate a simple model of the important factors you think directly affect the subject. The factors affecting species abundance and distribution fall into two broad categories: abiotic and biotic causes or variables. Abiotic factors are any variable in the environment that is not living. These abiotic factors include, but are not limited to, light intensity, temperature, variation in temperature, length of growing season, fire regimes, soil moisture, soil nutrient availability, rain fall, and seasonal variation in rain fall. Biotic factors are any variable in the environment that is created by another living organism. Biotic factors include, but are not limited to, competition, herbivory, mutualism, and disease. The basis for your research hypotheses will be about relationships between 2 or more of these variables and individuals, populations, or communities.
The process of forming a clear hypothesis and devising acceptable tests is called "strong inference". It is described in a famous paper by Platt. In the paper, Strong Inference, Platt also describes why you should devise alternative hypotheses.
Try developing several questions related to your topic, and then choose one that best fits. Rephrase your question until it is both analytical, requiring analysis of data, and focused and based upon previous information on the topic to prevent a superficial discussion. How do ecological theories help shape your hypothesis?
2- Developing an experimental design. How are ecology experiments designed? Your research hypothesis, experiment, and consequently your results have to be related.
Why do we sample the environment? The goal in science is often to try to estimate some parameter of interest. It's impossible to measure every individual of a population of interest. Generally, the more samples you collect, the more confident you can be about the results you find. By taking measurements in several or many study plots (replicating), and averaging your results, you can account for the effect of natural variation (heterogeneity) present in the forest. This natural variation of the environment can be very misleading if you are not aware of it. In general, we will want as large of a sample as possible. This will improve our confidence that the estimate we generate from the sample is close to the actual parameter we are interested in knowing.
It is impossible in nature to find two identical locations for an experiment. If you are conducting an experiment, to avoid the error of overlooking a variation, you need to assign the "treatment" and the "control" randomly. What problems can arise when you design and carry out your experiment? Did you have enough samples? Two or more representatives of each treatment, called replications, are necessary to ensure that you obtain a sufficient sample size and a representative sample size.
In his paper, Pseudoreplication and the design of ecological field experiments, Stuart H. Hurlbert reviews 176 experiments and finds that pseudoreplication occurred in 27% of them.
3- Obtaining evidence by observation and measurement. Your observations must be somehow accountable.
4- Using logic and insight to analyze data. Data analysis is about extracting meaningful information from your dataset using an appropriate tool. Displaying the data in graphical form often helps you to extract meaningful information. However, there is often a great deal of stochasticity (randomness, or noise) in ecological systems, so ecologists often employ tools whose specialty is separating meaning from noise - statistics! How does the use of evidence (data collected and analyzed) help to confirm or disprove your hypothesis? Your qualitative conceptual model may help you understand what your results mean also. Click here to see an example.
5- Interpretation. Why must you coordinate ecological theory and your evidence to develop an explanation for your findings?
Your experiment is an evaluation of how likely your explanation of the pattern or phenomenon is. Therefore, you need to carefully explain how your results mesh with , or contradicts, your previous explanation of ecological theory, and how you can explain the differences.