USING RESOURCE UTILIZATION FUNCTIONS (RUFs) TO ASSESS WILDLIFE-HABITAT RELATIONSHIPS Y = β 0 + β 1 x 1 + β 2 x 2 + β 3 x 3 + β n x n …… Brian Kertson Wildlife.

Slides:



Advertisements
Similar presentations
Regression Analysis Once a linear relationship is defined, the independent variable can be used to forecast the dependent variable. Y ^ = bo + bX bo is.
Advertisements

Predictive Model of Mountain Goat Summer Habitat Suitability in Glacier National Park, Montana, USA Don White, Jr. 1 and Steve Gniadek 2 1 University of.
Simple Linear Regression and Correlation
Multiple Regression [ Cross-Sectional Data ]
Chapter 13 Multiple Regression
Spring INTRODUCTION There exists a lot of methods used for identifying high risk locations or sites that experience more crashes than one would.
Regression Analysis. Unscheduled Maintenance Issue: l 36 flight squadrons l Each experiences unscheduled maintenance actions (UMAs) l UMAs costs $1000.
PPA 501 – Analytical Methods in Administration Lecture 8 – Linear Regression and Correlation.
Chapter 12 Multiple Regression
Statistics for Business and Economics
Multiple Regression Models. The Multiple Regression Model The relationship between one dependent & two or more independent variables is a linear function.
Evaluating Animal Space Use: New Developments to Estimate Animal Movements, Home Range, and Habitat Selection Dr. Jon S. Horne, University of Idaho.
Nemours Biomedical Research Statistics April 2, 2009 Tim Bunnell, Ph.D. & Jobayer Hossain, Ph.D. Nemours Bioinformatics Core Facility.
Lecture 24: Thurs. Dec. 4 Extra sum of squares F-tests (10.3) R-squared statistic (10.4.1) Residual plots (11.2) Influential observations (11.3,
Basic Business Statistics, 11e © 2009 Prentice-Hall, Inc. Chap 14-1 Chapter 14 Introduction to Multiple Regression Basic Business Statistics 11 th Edition.
Influence of variation in individuals, and spatial and temporal variation in resource availability on population dynamics. Jalene M. LaMontagne.
1 Simple Linear Regression Chapter Introduction In this chapter we examine the relationship among interval variables via a mathematical equation.
Chapter 11 Multiple Regression.
Mahanalobis Distance Dr. A.K.M. Saiful Islam Source:
Ch. 14: The Multiple Regression Model building
Dr. Mario MazzocchiResearch Methods & Data Analysis1 Correlation and regression analysis Week 8 Research Methods & Data Analysis.
BCOR 1020 Business Statistics Lecture 24 – April 17, 2008.
Today Concepts underlying inferential statistics
Correlation and Regression Analysis
Correlation Coefficients Pearson’s Product Moment Correlation Coefficient  interval or ratio data only What about ordinal data?
McGraw-Hill/Irwin Copyright © 2013 by The McGraw-Hill Companies, Inc. All rights reserved. A PowerPoint Presentation Package to Accompany Applied Statistics.
Introduction to Regression Analysis, Chapter 13,
So are how the computer determines the size of the intercept and the slope respectively in an OLS regression The OLS equations give a nice, clear intuitive.
Chapter 12 Inferential Statistics Gay, Mills, and Airasian
Marshall University School of Medicine Department of Biochemistry and Microbiology BMS 617 Lecture 12: Multiple and Logistic Regression Marshall University.
Regression and Correlation Methods Judy Zhong Ph.D.
Introduction to Linear Regression and Correlation Analysis
Inference for regression - Simple linear regression
Simple Linear Regression
Environmental Modeling Steven I. Gordon Ohio Supercomputer Center June, 2004.
Statistics for Business and Economics Chapter 10 Simple Linear Regression.
© 2001 Prentice-Hall, Inc. Statistics for Business and Economics Simple Linear Regression Chapter 10.
Correlation and Regression Used when we are interested in the relationship between two variables. NOT the differences between means or medians of different.
EQT 373 Chapter 3 Simple Linear Regression. EQT 373 Learning Objectives In this chapter, you learn: How to use regression analysis to predict the value.
NR505 GIS Applications in Wildlife Sciences
Lecture 8 Simple Linear Regression (cont.). Section Objectives: Statistical model for linear regression Data for simple linear regression Estimation.
Inference for Regression Simple Linear Regression IPS Chapter 10.1 © 2009 W.H. Freeman and Company.
Extent and Mask Extent of original data Extent of analysis area Mask – areas of interest Remember all rasters are rectangles.
Discrete Choice Modeling William Greene Stern School of Business New York University.
Spatial Interpolation Chapter 13. Introduction Land surface in Chapter 13 Land surface in Chapter 13 Also a non-existing surface, but visualized as a.
Chapter 16 Data Analysis: Testing for Associations.
Business Statistics: A Decision-Making Approach, 6e © 2005 Prentice-Hall, Inc. Chap 13-1 Introduction to Regression Analysis Regression analysis is used.
COUGAR ECOLOGY AND BEHAVIOR IN AN INCREASINGLY URBAN WORLD Brian Kertson Wildlife Science Group WACFWRU/SFR University of Washington.
Chapter 6 Simple Regression Introduction Fundamental questions – Is there a relationship between two random variables and how strong is it? – Can.
Lecture 10: Correlation and Regression Model.
BIOL 4240 Field Ecology. Ecologists are often interested in spatial data… Plant ecologists, distribution of individuals. Animal ecologists, distribution.
Chapter 22: Building Multiple Regression Models Generalization of univariate linear regression models. One unit of data with a value of dependent variable.
28. Multiple regression The Practice of Statistics in the Life Sciences Second Edition.
1.Define a landscape. What is the focus of Landscape Ecology. Notes 2. Discuss the role of spatial and temporal scale in affecting landscape composition,
Basic Business Statistics, 10e © 2006 Prentice-Hall, Inc.. Chap 14-1 Chapter 14 Introduction to Multiple Regression Basic Business Statistics 10 th Edition.
Phil Hurvitz Avian Conservation Lab Meeting 8. March. 2002
Introduction to Multiple Regression Lecture 11. The Multiple Regression Model Idea: Examine the linear relationship between 1 dependent (Y) & 2 or more.
Why use landscape models?  Models allow us to generate and test hypotheses on systems Collect data, construct model based on assumptions, observe behavior.
Multiple Regression Learning Objectives n Explain the Linear Multiple Regression Model n Interpret Linear Multiple Regression Computer Output n Test.
Biostatistics Regression and Correlation Methods Class #10 April 4, 2000.
Educational Research Inferential Statistics Chapter th Chapter 12- 8th Gay and Airasian.
7. Air Quality Modeling Laboratory: individual processes Field: system observations Numerical Models: Enable description of complex, interacting, often.
Use & Availability of Habitats & Foods Resource selection Measurements of use & availability –Food –Habitat Design & analysis Modeling Sampling.
Inference about the slope parameter and correlation
Chapter 14 Introduction to Multiple Regression
Multiple Regression Analysis and Model Building
CHAPTER 29: Multiple Regression*
Landscape Connectivity and Permeability
Introduction to Regression
Nazmus Saquib, PhD Head of Research Sulaiman AlRajhi Colleges
Presentation transcript:

USING RESOURCE UTILIZATION FUNCTIONS (RUFs) TO ASSESS WILDLIFE-HABITAT RELATIONSHIPS Y = β 0 + β 1 x 1 + β 2 x 2 + β 3 x 3 + β n x n …… Brian Kertson Wildlife Science Group SFR/WACFWRU

HABITAT IS THE KEY FOR WILDLIFE Understanding relationships is critical -Food -Reproduction -Survivorship -Predator-prey dynamics -Behavior and ecology Management and conservation

KEY TERMINOLOGY Use: how much, how often – metric matters Selection/Avoidance: animal uses resource more or less than available Preference: animal selects between 2 equally available resources

WILDLIFE-HABITAT METHODS Many analytical procedures available Common techniques: -Compositional Analysis -Resource Selection Functions (RSFs) -Resource Selection Probability Functions (RSPFs) Varying degrees of rigor, each has advantages and disadvantages

COMMON PROBLEMS Lack of independence of observations Incorrect sampling unit Habitat data and scale -Use of remote sensing Unit-sum constraint Discrete use Failure to connect with behavior (i.e., fitness)

USED VS. UNUSED LIMIATIONS Logistic regression Contamination: -Classified as unused when it was used -GPS -Snow tracking -Critter cams

PROBLEMS DEFINING AVAILABILITY Can we know how animals perceive their environment? Do we actually know what is available? NO!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! Arbitrary Home range simulations: -Rigorous: potentially -Biologically meaningless You know nothing!! Stupid hairless monkeys.

ADDITIONAL AVAILABILITY ISSUES Kertson and Marzluff, in press

Resource Utilization Functions Marzluff et al (Ecology) Continuous: -High vs. low use (relative comparison) Multivariate: -Multiple regression Individual is sampling unit: -Quantify individual variation No measure of availability

HOW THE RUF WORKS Animal relocations 99% Utilization Distribution (Use values) Sampling grid Use and habitat covariates Ruf.fit

KEY TOOLS ArcMap 9.x Hawth’s Tools: -Bivariate kernel Excel or Notepad R statistical computing -Ruf.fit package

MEASURING USE Individual = sampling unit Sampling design critical -Individuals -Monitoring Increase monitoring, more refined UD VHF vs. GPS: -Increased resolution -Increased accuracy -Not perfect Kertson and Marzluff, in press

UTILIZATION DISTRIBUTION (UD) Animal use is not random -Gradient of use Probability Density Function (pdf) -Sums to 1 Use = height (volume) of UD

UD ESTIMATION Fixed kernel Min. of 30 relocations -Preferably n ≥ 50 Resolution (grid size): -25 or 30 m common Bandwidth smoothing (h) -Most critical component

SELECTION OF h Selection: data Over vs. under-smoothing Univariate vs. bivariate Lots of options: -Reference (HREF) -Least-squares cross-validation (LSCV) -Plug in (PI) -Solve the equation (STE) -Biased cross validation (BCV) Each has +/-

EFFECTS OF h ON UD Kertson and Marzluff, in press

ESTIMATING h Animal Movements Extension (ArcView 3.3) ArcMap 9.x: -Home Range Tools (HRT) *LSCV, BCV, HREF R statistical computing: -KernSmooth package -KS package *PI values from both (bivariate)

UD CHALLENGES UD size can push the limits of software: -Male cougar UDs can exceed 2.0 million points Over-smoothing: -Lakes, rivers, major highways, and other unsuitable/unusable habitat Under-smoothing: -Donut holes and disconnect cores Solutions: -Clip UD (over-smoothing) -Adjust h (try different bandwidth method) -Little bit of black magic here

LANDSCAPE COVARIATES Covariate types: -Categorical -Continuous Resolution: -As fine as possible -Landscape configuration -Remote sensing Transformations: -Can improve model performance Distance to Water Percent Conifer Forest

CATEGORICAL COVARIATES Common categorical covariates: -Landcover -Aspect Classes for each variable are not independent Must be recoded 0,1 -No. of columns = no. of classes

CATEGORICAL COVARIATES LC_5Conifer Forest Mixed Forest RiparianHigh Elevation Urban

RUF.FIT Developed by Dr. Mark Handcock (UW-CSSS) Multiple regression: Y = β 0 + β 1 x 1 + β 2 x 2 + β 3 x 3 + β n x n …… Code: Cat1<- ruf.fit(USE ~ COV1 + COV2 + COV3 + COV4, space= ~ X + Y, data=data_file, theta=hval, name=“whatever_you_want", standardized=F) Corrects for spatial dependence in UD Unstandardized and standardized coefficients

MODEL COEFFICIENTS Average for sample Coefficient signs: -Increase use (+); decrease use (-) Unstandardized: -Mapping predicted occurrence Standardized: -Statistical significance of individual covariates -Differences between covariates -Relative importance -Proportion of sample +/- influence

RUF.FIT OUTPUT > summary(CAT1) Standardized Coefficients for name: Misska Matern Log-Lik = LS Log-Lik = Change in Log-Lik p-value = < 1e-04 MLE s.e. LS estimate LS s.e. range NA NA smoothness NA NA NA (Intercept) PCCREG PCF PFOREST DWATER DISTEDGE DISTROAD DRESD RESDENS1KM PAR SLOPE DEM β and associated SE

HOW DOES LANDSCAPE INFLUENCE COUGAR-HUMAN INTERACTION? Apex predator with a large home range Largest geographic distribution of any terrestrial mammal in western hemisphere -Tremendous habitat diversity Key landscape resources: -Ungulate prey -Cover High levels of interaction with people

METHODOLOGY Captured 32 cougars in western WA, outfitted with GPS collars Investigated interaction reports Focused on landscape metrics I suspect correlate with prey and cover Modeled with RUF Quantified individual variation

UNSTANDARDIZED COEFFICIENTS

COUGAR PREDICTED USE

STANDARDIZED COEFFICIENTS

CONSERVATION AND MANAGEMENT IMPLICATIONS Identify key resources to manage and conserve Identify high quality habitats Develop proactive management strategies -71.5% of confirmed interactions occurred in high and med-high use habitats -Management hotspots Space use and interactions with people highly individualized -Population approaches may not work

REGIONAL APPLICATIONS

ADDITIONAL APPROACHES Sex and age specific RUFS: -Male vs. female -Adult vs. subadult Behavior specific: -Movement rates -Relates habitat use to different aspects of fitness Traveling Resting/Feeding Hunting Nursing

RUF CHALLENGES The more RAM the better Capable of running full data set, may need to sub- sample Processing time can be significant Model comparisons (e.g., model parsimony) difficult -RUF outputs log-likelihoods (ΔAIC)

RUF LIMITATIONS Models are tools, not absolute truth Results are only as good as the data used -Limitations and accuracy of remotely-sensed data Do the results pass the laugh test? Subject to same assumptions as normal multiple regression No alternatives for correcting spatial dependence in UD

NEED DIRECTIONS?