Are Peromyscus populations shifting in response to climate change?

The Context and Question

Species ranges depend on interactions between their traits and the physical environment, and often on species interactions such as competition. Climate change is altering species ranges through its effects these interactions. Jeremy Papuga and Dr. Susan Hoffman are examing those effects in two species of Peromyscus, P. maniculatus gracilis and P. leucopus n.. They have anecdotal evidence that the two species are moving north and that P. leucopus is replacing P. maniculatus in southern sites. Whether they are competing, or whether each is responding to other factors is a different story.

Map of Michigan study sites

Jeremy’s data include numbers of juvenile and adult mice caught in monthly trappings in six sites (see map) over four years. It also includes information about weather, including,

• HDD - “heating degree-days” are the cumulative degrees F below 65 deg F, from Dec.-March each year. For instance, two consecutive days at 55 deg F would sum to 20 (10 + 10) HDD.
• ave.min - the average daily minimum temp from Dec-March (deg F).
• warm.up.day - the first day of the spring when there were five consecutive days above freezing. The units are Julian or ordinal day, which is the number of days following January 1. All of the days in these data are between May 1-20.
• A_PLN and A_PMG are the numbers of adults live trapped at each site in each May.
• J_PLN and J_PMG are the numbers of juveniles live trapped at each site in each May.
• Site - refers to the study site (refer to map).

Your goal

1. Brainstorm causal relationships regarding the various sites and weather, juvenile and adult population sizes of the two species, and adult population growth rates.
2. Put your question or proposed hypothesis in words.
3. Make a quantitative prediction that follows from your hypothesis and the available data.
4. Create a picture of your prediction - sketch a graph of your prediction.
5. Evaluate evidence for one causal relationship by (i) graphing the data in the same manner as you proposed in your prediction, and (ii) describing what the data say about your hypothesis or answer your question.
6. Turn in (a) your original question, (b) the picture of your prediction, (c) the graph of data that assess your prediction, and (d) your description (paragraph) of what the data say about your hypothesis or answer your question.

The data

I have created three different data sets that arrange the data in three different ways. Download them to your R work directory and then load them into R to view them.

Import the data sets.

## wide format of weather and population sizes
MayN.wide <- read.csv("MayN_wide.csv")
## peek at the structure of a data set
View(MayN.wide)

## long format of weather and population sizes
MayN.long <- read.csv("MayN_long.csv")
View(MayN.long)


## long format of weather and per capita population growth rates
Mayr.long <- read.csv("Mayr_long.csv")
View(Mayr.long)

Graph your data. Follow these instructions to use ggplot

### FIRST LOAD THE GGPLOT2 LIBRARY
library(ggplot2)

• ggplot() defines the data set, the x-variable, the y-variable, and any grouping variable.
• adding geom_point() creates a scatterplot.
• adding geom_smooth( ) adds a fitted line, either one line or a line for each group if you identified a grouping variable. You can add a method that specifies what kind of fit you want.
• geom_line() connects each point to adjacent points (a “line graph”).
• …and there are many, many other types of geom_

Here is an example of a scatterplot with a simple line connecting observations.

ggplot(data=MayN.wide, aes(x=Year, y=A_PLN, colour=Site)) + 
    geom_point() +
    geom_line() + 
    labs(x="Calendar year", y="Number of adults") 

To save this graph to your working directory, use ggsave(), for instance,

ggsave("myGraph.png", width=5, height=6) # units are inches

The file extension (png or jpg or pdf) determines the file type.

In sum: Think about causal relationships; represent the predictions in a quantitive graph; graph the actual data; evaluate your hypothesized causal relationship.