Case Study

Distance sampling to estimate the population density of magpies and the factors affecting their distribution. This project will require you to develop a hypothesis, develop a sampling design appropriate to test it, collect and analyze the data.

Reading

  1. CWP CH 2
  2. CWP CH 4
  3. Christianson & Creel 2014. Ecosystem Scale Declines in Elk Recruitment and Population Growth with Wolf Colonization: A Before-After-Control-Impact Approach. Plos One 9(7): e102330 - discussed in class as introduction to estimation of population size and effects on it: wolf counts, elk counts and inferences.
  4. Chandler, R. 2014. Distance sampling analysis in unmarked.
  5. ABGR CH 3
  6. ABGR CH 5 (skip CH 4 for now)
  7. LIon example of CMR estimation of population size and survival rate.

R code examples and exercises

  1. OLS Regression model example used in class
  2. A short summary of regression
  3. R: Exercise Three A  Hypothesis testing with simple and multiple regression
    1. Data file: kenyaherdsize2.txt
  4. R exercise Three B: Generalized Linear Models.
  5. R: Exercise Three C Model selection and multimodel inference
    1. Data file: kenyaherdsize3.txt
  6. R: Exercise Four Estimating population density with distance sampling.  The R script as a .R file, rather than html , up to the point of fitting a distance sampling model, to save you cutting and pasting code into the script editor:
    1. Data files:
      1. knp.allsp.oct2012.csv
      2. knp_covs_oct2012.csv
  7. BIOE 521 only:  R Exercise 5: Estimating survival rate with CJS models  using RMark.

Homework

HW 2A is due in class WEDNESDAY 10/2 (ignore the old due date on the link). Scroll all th ay down the page for HW and dataset links.

HW 2B is due in class MONDAY 10/7  (ignore the old due date on the link). Scroll all the way down the page for link.

HW3 (completed datasheet) is due in class MONDAY 10/14A few things to remember: (1) Each flock (not each bird) is one detection (one row of datasheet).  (2) Covariates are measured at the level of transects (not flocks). (3) Don't forget to record the length of each transect (spaces provided at top for 10 transects, add as needed).  (4) Aim for 30 detections - more if your model is complex.

HW4 (distance sampling analysis) is due in class MONDAY 10/21.  Resources to help you with this are: (1) R Excercise 4 above with example code run on a dataset for puku, (2) Detailed step-by-step instructions (item 4 below) for the process of fitting a distance sampling model in unmarked, (3) Details on how to do the initial steps of converting the raw data to sighing angles between 0 and 90 degrees, and conversion of degrees to radians before calculating sin(sighting angle), which you need to calculate perpendicular distances. (4) Explanation (at the very bottom, below) on how to deal with transects that no magpie sightings, if that occurred in your data.

  1. HW2A - OLS and GLM regression in R.
    • Data sets for HW2A:  
    • hwq1data 
    • hwq2data
    • Remember that there is a menu option in R Studio to set the working directory:
      • Create a new R script to do the analysis: under the File tab on the top menu, select 'New File' then click 'R Script' on the pop-up menu.
      • Save your R script to the location where you put the datafiles when you downloaded them (for example, R script file and datafile both on the desktop).
      • Under the Session tab on the top menu in R Studio, select 'Set Working Directory' and click 'To Source File Location'.
      • Read a data file into R with a line of code such as: data.for.q1 <- read.csv("hwq1data")
  2. HW2B - hypothesis statement, sampling design and methods for magpie data collection.  
  3. HW3 - magpie data  Example data sheet.  Copy of completed data
  4. HW 4 - Detailed instructions for magpie analysis