README.md

encounteR

Build Status codecov.io

A package for estimating population densities and other ecological measurements using Random Encounter Models.

Random encounter models provide an alternative method for estimating animal densities, contact rates and various other measures important in ecology. They can replace distance sampling methods and mark-recapture methods and have potentially less restrictive assumptions. Notably, one does not need to measure the distance between your sensor (camera trap, acoustic sensor etc.) and the animal, nor do you need individual recognition of animals.

This package is an attempt to pull together different models and functions useful for REM style analyses into one package.

Please feel free to get in touch, mention models that should be here that aren't or contribute code.

Installation

library(devtools)
install_github('timcdlucas/encounteR')
library(encounteR)

Web App

If you do not use R, a very basic web app is available. It implements the gREM model (of which the REM model and gas models are special cases) but nothing else.

Road map

Includes: - gREM model - REM model - Gas model

Could include: - 3D models - Acoustic parameter estimation for gREM - Varied environments - Moving detectors

References

Generalised Random Encounter Model:

Lucas, T. C. D., Moorcroft, E. A., Freeman, R., Rowcliffe, J. M., Jones, K. E. (2015), A generalised random encounter model for estimating animal density with remote sensor data. Methods in Ecology and Evolution. doi: 10.1111/2041-210X.12346

Random Encounter Model:

Rowcliffe, J. M., Field, J., Turvey, S. T. and Carbone, C. (2008), Estimating animal density using camera traps without the need for individual recognition. Journal of Applied Ecology, 45: 1228–1236. doi: 10.1111/j.1365-2664.2008.01473.x

Gas Model:

Yapp, W. B. "The theory of line transects." Bird study 3.2 (1956): 93-104. APA PDF

Gas model review:

Hutchinson, J. M. C. and Waser, P. M. (2007), Use, misuse and extensions of “ideal gas” models of animal encounter. Biological Reviews, 82: 335–359. doi: 10.1111/j.1469-185X.2007.00014.x



timcdlucas/encounteR documentation built on May 31, 2019, 1:48 p.m.