Method draft of a biology research paper

Items that are due:

A draft of your methods section
A “habitat” variable in the traits spreadsheet, describing the site where you made personal leaf collections as either “mesic” or “xeric” We are going to try adding this variable into our linear models to explain more of the variation in the relationship between stomata and MAP, in much the same way Olson et al. 2014 used “habit” as a variable to improve their model relating vessel diameter to stem diameter.

Methods

The Methods section (aka Materials and Methods) of a scientific manuscript is where you describe exactly what you did to test your hypothesis. It typically includes descriptions of the study site (for field studies), the experimental design, and the statistical analyses used.

The main point of the Methods section is to provide the reader with enough detail that they could replicate your experiment, without overwhelming them with excessive detail that isn’t needed.

Important things to keep in mind when writing the Methods section:

You are writing as if for a scientific journal — therefore the readership is a general scientific audience and doesn’t need to know highly specific details about our class or where exactly we carried out the experiment. (For example, don’t say “Everyone in our class was assigned two GBIF species to measure for leaf traits and leaf area”, say something like “In addition to field-collected leaf samples, we also measured leaf traits and leaf area for 52 species for which specimen images were obtained from the Global Biodiversity Information Facility (GBIF) database (www.gbif.org).”)
Do not list materials. Many students have this habit as a result of years of following lab manual protocols. Instead, mention materials in the text as you describe how they were used. Avoid any kind of bullet point-based or numbered list.
Do not mention recording data or entering it into a spreadsheet – these are basic tasks anyone with a scientific background if familiar with, and don’t need to be described explicitly.
Avoid excessive use of passive voice. In the past, most scientists thought passive voice was preferable because they thought it sounded more objective. The problem with writing excessively in the passive voice (“…the leaves were collected at such-and-such locality…” or “…the data were analyzed using the statistical software R…”) makes the subject of the sentence invisible, which in turn makes the reader have to work harder to understand what the text is saying. Some amount of passive voice is okay (I would say keep it under ½ of the sentences), but be careful not to slip into all passive voice all the time.
The methods section should be written in first person (“I” or “we”), past tense. In other words, it is a detailed description of work you have already done.
Brand names of instruments and equipment like balances, glassware, etc are not needed, unless the equipment is highly specialized (e.g., a PCR thermocycler, a leaf photosynthesis analyzer).
Make sure to describe replication and statistical tests performed (use the command “dim()” on the data frame in R to get the number of rows and columns. Number of rows is the number of replicates).
If you use an abbreviation like CLAMP, write it out the first time it appears in the text, with the first letter of each word capitalized, followed by the abbreviation in parentheses. After this you can just use the abbreviation:
Global Biodiversity Information Facility (GBIF), after which you can use GBIF
Climate Leaf Analysis Multivariate Program (CLAMP), after which you can use CLAMP
For websites, give the web address in parentheses after the first time you mention the site. If you are mentioning both a website and an abbreviation, put both the abbreviation and web address in parentheses, separated by a semicolon.
Global Biodiversity Information Facility (GBIF; www.gbif.org) (Or see how I did it in the first bullet point above.)
To find the “right” amount of detail, start by assuming the reader has a general science background, so won’t need to be told how to do techniques that are generally well known. You usually won’t need to describe specific actions taken, such as placing a solution into a beaker or entering data into a spreadsheet — this is an example of “too much” detail. Here are some more examples illustrating too much, too little, and the right amount of detail:

Too much Just right
When we found our assigned plot, one person stood right where the point was and held onto the end of a three meter long rope. Another person held the other end and made a circle in the dirt using a stick. We sampled in a circular plot within a 3 m radius.
We divided the class into two groups and wrote each person’s name down in an Excel spreadsheet. We then lined up in the hall outside the classroom where we set the measuring tape. Each student jumped twice: once with a running jump and once from a standstill. (This is actually both too much detail and too little because it leaves out some important information — see “Just right” column.) Each subject jumped twice: once with a running jump and once from a standstill. To avoid any confounding effects related to jump order, we had half of the subjects do the running jump first and the other half do the standing jump first. We measured jump distance at the landing point of the heel.

Not enough Just right
We blended our objects together with water and poured each blended object individually into separate containers. We then chose a single solution to begin testing. We dipped a small piece of filter paper into the blended solution of our choice and then dropped it into a small container of hydrogen peroxide. We timed it from the point of release and until it came back to the surface. We repeated this process five times with each solution. To test this hypothesis, we blended 10g of banana fruit with 100 mL of water plus a handful of ice. We then strained the resulting mixture and added water to bring the strained solution to a final volume of 100 mL. This was repeated for the other 3 substances: the banana stem, leaves, and peel. We placed all 4 solutions on ice to ensure the catalase stayed fresh. We prepared a 50-mL beaker of hydrogen peroxide and discs of filter paper produced using a standard hole-punch. We then soaked a filter paper piece in solution for 5 seconds, took it out and dabbed it on a paper towel, then dropped it in the hydrogen peroxide and timed how long it took to float to the top. We repeated this 5 times per solution. The enzyme activity rate was calculated by dividing the height of hydrogen peroxide in the small beaker by the time the filter paper took in the beaker (mm/s). We analyzed the data using an ANOVA statistical test.

Here are the key pieces of information for our project to make sure you include in your methods draft. You may use this text as a skeleton, but be sure to modify it as appropriate depending on the variables you tested, and also to make it flow better. Only describe the parts of the methods relevant to your hypotheses: if you didn’t do anything with stomatal density, for example, then you don’t need to describe the stomatal cast part of the methods.

Each of 23 students in the class collected leaves from two sites. At each site they collected five leaves from each of up to ten species.
We made leaf casts using clear nail polish to measure stomatal density. Nail polish peels were measured under a compound light microscope at 100x or 400x magnification. (Don’t need to describe how density was calculated — this is something that would be pretty standard for anyone with a science background.)
Leaves were then mounted on a white paper background and scanned as .jpg files with a ruler for scale.
Each leaf was analyzed for leaf functional traits. We scored leaf traits following the CLAMP protocol (give the website) for the following traits: lobing, teeth, regularity of tooth spacing, closeness of teeth, teeth rounded or appressed, teeth acute, teeth compound, apex form, length-to-width ratio, and leaf shape. (Only describe the traits you used in your analysis. Give more detail on what the trait means and how it corresponds to the numerical values we gave them — for example, if you analyzed leaf shape, describe what the shapes were and what the number between 0 and 1 actually means for leaf shape. This is very important because the indices used in CLAMP don’t have any obvious meaning unless you are familiar with their scoring).
We measured leaf lamina area using the ImageJ analysis software (include their website) using either manual tracing (for leaves that were not clearly separated against the white background) or automatic tracing using the magic wand tool (for leaves that were isolated against the background). (Again, don’t need to go into a ton of detail about the exact protocol because this information is widely available).
To augment the geographical range of our dataset, each student also analyzed two additional western North American species using specimen images obtained from GBIF. Specimens had to have geographic coordinates and images. GBIF specimens were analyzed for leaf traits and leaf area only.
For each locality in the dataset, we extracted data on mean annual temperature (MAT) and mean annual precipitation (MAP) using the climate WNA website (give web address).
Linear regression in the statistical software R 4.0.3 (to get the citation for R, type “citation()” in the console) to test the relationship between [leaf traits/area/stomatal density] and [MAT/MAP] — modify this text based on the three leaf-climate relationships you chose to test.
Give sample sizes (this will be the number of rows in the stomata.mean, area.mean, and/or traits.mean data frames. You can find this by using the command dim(data.frame) in R. You can’t get it from looking at the .csv spreadsheets in Excel because for all three we took averages of measurements, which made them smaller.