University of Southern California Dissertations and Theses

Using Landscape Integrity Index to evaluate the cumulative impacts of BLM resource management programs

(USC Thesis Other)

Using Landscape Integrity Index to evaluate the cumulative impacts of BLM resource management programs

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

Using Landscape Integrity Index to Evaluate the Cumulative Impacts of
BLM Resource Management Programs

by

Liling Lee

A Thesis Presented to the
Faculty of the USC Dornsife College of Letters, Arts and Sciences
University of Southern California
In Partial Fulfillment of the
Requirements for the Degree
Master of Science
(Geographic Information Science and Technology)

May 2020

Copyright © 2020 by Liling Lee

To my parents, Ngawong Ngoon and Yuhain Lee,
for always reminding me to eat and sleep regularly,
and my grandma, Jui Chu Lee,
for constantly cheering me on this journey

iv
Table of Contents
List of Figures ............................................................................................................................... vii
List of Tables ............................................................................................................................... viii
Acknowledgments.......................................................................................................................... ix
List of Abbreviations ...................................................................................................................... x
Abstract ......................................................................................................................................... xii
Chapter 1 Introduction .................................................................................................................... 1
1.1. Study Area ..........................................................................................................................3
1.2. Motivation ...........................................................................................................................4
1.2.1. A Comprehensive, Standardized, and Transparent GIS Approach ............................4
1.2.2. Evaluation of Ecological Integrity in CEA ................................................................4
1.2.3. Landscape Metrics .....................................................................................................5
1.2.4. Multiscale Data ..........................................................................................................6
1.2.5. Court Challenges ........................................................................................................7
1.3. Research Purpose, Goals, and Objectives ...........................................................................7
1.4. Research Questions .............................................................................................................9
1.5. Thesis Structure ..................................................................................................................9
Chapter 2 Related Work................................................................................................................ 11
2.1. BLM’s Mission .................................................................................................................11
2.1.1. Landscape Approach ................................................................................................12
2.2. Cumulative Effects Analysis.............................................................................................13
2.2.1. History of Cumulative Effects Analysis ..................................................................13
2.2.2. Defining Cumulative Effects ...................................................................................13
2.2.3. Evaluating Cumulative Effects ................................................................................15
2.3. Challenges to Cumulative Effects Analysis ......................................................................16
v
2.3.1. Spatial and Temporal Scale .....................................................................................16
2.3.2. Indirect and Future Effects .......................................................................................17
2.4. Ecological Integrity and Its Indicators ..............................................................................18
2.4.1. Composition, Structure, and Function .....................................................................18
2.4.2. BLM Management Indicator Species ......................................................................19
2.5. Use of Landscape Metrics .................................................................................................20
2.6. Landscape Integrity Index.................................................................................................22
Chapter 3 Methodology ................................................................................................................ 24
3.1. Study Area ........................................................................................................................24
3.2. Data Sources and Selection of Indicators & Metrics ........................................................25
3.2.1. Ecological Integrity Indicators .................................................................................26
3.2.2. Resource- and Stressor-based Metrics .....................................................................28
3.2.3. Landscape Metrics ...................................................................................................28
3.3. Three-Stage Workflow ......................................................................................................32
3.3.1. Understand management context .............................................................................33
3.3.2. Design assessment ...................................................................................................35
3.3.3. Implement assessment .............................................................................................36
3.4. Procedure ..........................................................................................................................38
3.4.1. Data Preparation.......................................................................................................39
3.4.2. Landscape Metrics in R ...........................................................................................39
3.4.3. Composite Scoring System ......................................................................................40
3.4.4. Moving Window Analysis .......................................................................................45
3.4.5. Model Validation .....................................................................................................45
Chapter 4 Results .......................................................................................................................... 48
4.1. The Landscape Integrity Index Model ..............................................................................48
vi
4.2. Landscape Diversity Metrics ............................................................................................51
4.3. Model Validation Results .................................................................................................53
Chapter 5 Discussion and Conclusions ......................................................................................... 58
5.1. Landscape Integrity of CFO Planning Area......................................................................58
5.2. Applications of LII ............................................................................................................59
5.3. Research Limitations ........................................................................................................60
5.4. Future Research Opportunities .........................................................................................62
5.5. Conclusion ........................................................................................................................64
References ..................................................................................................................................... 66
Appendix A Management Indicator Species for Carlsbad Field Office, NM............................... 71
Appendix B Landscape Metrics Results ....................................................................................... 73
Appendix C LII values and LCM Values for Model Validation .................................................. 90

vii
List of Figures
Figure 1 Map of BLM Carlsbad Field Office Planning Area .................................................... 3

Figure 2 Landscape Integrity Index (LII) Framework ............................................................... 8

Figure 3 (a) Relationships between activities (from resource
management programs), stressors, and ecological components;
(b) relationships between impact characterization and activity,
stressor, and/or ecological components. Source: Foley et al. (2017) ........................ 14
Figure 4 3-Stage Workflow to Assess Ecological Integrity.
Source: Carter et al. (2019) ........................................................................................ 33

Figure 5 Landscape Integrity Index (LII) Workflow ............................................................... 38

Figure 6 Histogram of LII values at Carlsbad Field Office ..................................................... 49

Figure 7 Map of Landscape Integrity Index at Carlsbad Field
Office Planning Area (2001-2018). LII values ranged from 0
(low landscape integrity) to 1 (high landscape integrity). ......................................... 50

Figure 8 Maps of Ecological Integrity, Landscape Metrics,
Resource-based Metrics, and Stressor-based Metrics at Carlsbad
Field Office Planning Area (2001-2018) ................................................................... 51

Figure 9 Scatter Plots from 2001, 2004, 2006, 2008, 2011, 2013,
and 2016 of (a) Shannon’s Diversity Index (SHDI); (b) Simpson’s
Diversity Index (SIDI); (c) Shannon’s Evenness Index (SHEI);
(d) Simpson’s Evenness Index (SIEI)........................................................................ 53

Figure 10 (a) Scatter Plot of LII and LCM Values; (b) Linear
Regression Line of LII and LCM Values; (c) Box Plot of LII and
LCM Values; (d) Density Plot of LII and LCM Values ............................................ 55

Figure 11 Landscape Integrity Index Map and Landscape Condition Map ............................... 56
Figure 12 Box Plot of LII Values within Protected Areas and Multiple-Use Areas.................. 56

viii
List of Tables
Table 1 Subject Area for Landscape Metrics, Definitions,
and Information Relevant to Landscape Fragmentation
or Landscape Diversity .............................................................................................. 21
Table 2 Spatial data inputs of management action and measure for
Ecological Indicators and Resource- and Stressor-based Metrics ............................. 25
Table 3 Landscape Metrics from FRAGSTATS and R with
Definition and Interpretation (2001, 2004, 2006, 2008, 2011,
2013, 2016) ................................................................................................................ 29
Table 4 Management Goals and Objectives of Vegetative
Communities and Minerals – Oil and Gas ................................................................. 34
Table 5 Site Impact Scores of Ecological Integrity Indicators ............................................... 42
Table 6 Site Impact Scores of Resource- and Stressor-based Metrics ................................... 43
Table 7 Model Validation Results for Linear Regression Model ........................................... 54
Table 8 Model Validation Results for Welch’s Two Sample T-Test ..................................... 57
Appendix A Management Indicator Species for Carlsbad Field Office, NM ........................... 71
Appendix B Landscape Metrics Results ................................................................................... 73
Appendix C LII values and LCM Values for Model Validation ............................................... 90

ix
Acknowledgments
I would like to give a big thanks to Professors Jennifer Bernstein and Karen Kemp for their
insightful guidance and constructive feedback on narrowing down my research proposal and
improving my thesis. I am grateful of my other faculty who offered much needed assistance in
advancing my writing and technical skills: Vanessa Osborne and Jennifer Swift. I greatly
appreciate my employer, Bureau of Land Management, who supported my work on the thesis
and provided excellent research ideas. A very special gratitude goes out to Dario Lunardi for
laying the cornerstone of this thesis and Calvin Deal for fostering a great thesis experience with
endless encouragement and invaluable field data. I would also like to thank the experts who
shared their knowledge with me during the making of this thesis: Tom Chatfield, Karla Rogers,
Devin Johnson, Alec Boyd, and Tony Stelle. Finally, this accomplishment would not have been
possible without my family and friends, who are always there for me with unwavering support
and unlimited trust in my potential. Thank you.

x
List of Abbreviations
AIM BLM Assessment, Inventory, and Monitoring Process
APD Application for Permit to Drill
BLM Bureau of Land Management
CEA Cumulative Effects Analysis
CEQ U.S. Council of Environmental Quality
CFO BLM Carlsbad Field Office
EIS Environmental Impact Statement
EVT Existing Vegetation Type
FLPMA Federal Land Policy and Management Act of 1976
GAP Gap Analysis Project
GIS Geographic information system
GISci Geographic information science
IEI Index of Ecological Integrity
IPA Important Plant Area
LANDFIRE Landscape Fire and Resource Management Planning Tools Project
LII Landscape Integrity Index
MRLC Multi-Resolution Land Characteristics (MRLC) Consortium
NEPA National Environmental Policy Act of 1969
NLCD National Land Cover Database
NM New Mexico
OHV Off-Highway Vehicle
RMP Resource Management
xi
SGCN Species of Greatest Conservation Need
SSI Spatial Sciences Institute
USC University of Southern California
USGS U.S. Geology Survey
VDEP Vegetation Departure

xii
Abstract
The Bureau of Land Management (BLM) is instrumental in connecting people with
public lands by providing and protecting opportunities to enjoy and use our country’s resources.
Understanding the cumulative effects of resource management programs is crucial for decision
makers to develop effective land management practices and appropriate allocation of funding
and resources. A comprehensive, standardized, and transparent GIS workflow can help visualize
and analyze ecological integrity, landscape patterns and processes, and promote a consistent
Cumulative Effects Analysis (CEA) and collaborative management across jurisdiction
boundaries.
This research evaluates the cumulative impacts of resource management programs in the
BLM Carlsbad Field Office (CFO), New Mexico by incorporating ecological integrity indicators,
resource- and stressor-based metrics, and landscape metrics to create a Landscape Integrity Index
(LII). Two resource management programs, Vegetative Communities and Minerals – Leasables –
Oil and Gas, were selected as the programs of interest for this study. The LII model considers the
management goals and objectives in the Draft BLM CFO Resource Management Plan (RMP) to
identify the necessary indicators and metrics. These indicators and metrics were each scored for
their site impact, distance decay function, or landscape metrics through the use of a Composite
Scoring System, and then combined into a single map. The resulting map with the LII values
shows areas of low landscape integrity near the urban and agricultural areas in CFO planning
area and high landscape integrity near central and southwest corner of CFO. CEA practitioners
and land managers will be able to address management goals and objectives, conduct a more
systematic and consistent analysis with relevant indicators and metrics, and visualize landscape
integrity using the LII framework.
1

Chapter 1 Introduction
As the largest land management agency in the nation, the Bureau of Land Management
(BLM) has a tremendous impact on how people interact with public lands with its dual
responsibility to manage public lands for multiple-use and conserve resources for the benefit of
present and future generations. The BLM is a federal agency within the Department of the
Interior that manages 246 million surface acres of public lands under the principles of multiple-
use and sustained yield. Multiple-use is defined as managing public lands and resource uses
collectively to best meet the needs of the present and future generations (U.S. Department of the
Interior 2001). Sustained yield is defined as the continuous high-level production of various
renewable natural resources via multiple-use in the public lands (U.S. Department of the Interior
2001). The BLM promotes multiple-use on public lands by supporting a variety of resource
management programs such as energy development, conservation stewardship, and recreation,
which can affect the ecological integrity of the lands in different ways. Ecological integrity refers
to the condition and ability of the ecological system to support biological communities with
abiotic components and provide ecological services (Hobbs et al. 2010, Wurtzebach and Schultz
2016, Carter et al. 2017).
Resource management programs can affect ecological integrity in both positive and
negative ways through surface disturbance, ecological function depletion or alteration, habitat
restoration, and other ecological processes. For example, energy developments including both
conventional (e.g. oil and gas, coal, or minerals) and renewable (e.g. wind, solar, and
geothermal) can disturb surface landscapes and negatively impact soil and water resources
(surface water and groundwater), wildlife habitat, and avian and bat species (Bureau of Land
Management 2018b). On the other hand, conservation stewardship activities such as vegetation
2

and noxious weed management and riparian and wetland actions promote long-term beneficial
impacts by restoring vegetation that meets ecological objectives, preventing soil erosion and
runoff, and ensuring water and vegetation quality (Bureau of Land Management 2018b). The
effects of a single resource management program are more straightforward to understand; but
when the area comprises multiple programs with potential contrasting management goals and
impacts, the cumulative effects of these programs are often difficult to assess.
Cumulative effects analysis (CEA), also referred to as cumulative impact assessment,
analyzes the cumulative effects of the actions on ecosystems. As defined by the Council of
Environmental Quality (CEQ), cumulative effects are “the impact on the environment which
results from the incremental impact of the action when added to other past, present, and
reasonably foreseeable future action”. In this study, these actions are from the resource
management programs of BLM Carlsbad Field Office, including the development of oil and gas
drilling wells and vegetative treatment, which affect the structure, function, or well-being of
various environments.
To better assess the cumulative effects of BLM resource management programs, my
research developed a Landscape Integrity Index (LII) for BLM Carlsbad Field Office, New
Mexico. The LII model serves as a landscape perspective of ecological integrity and land health,
incorporating ecological integrity indicators, resource- and stressor-based metrics, and landscape
metrics. Evaluating the cumulative impacts of different BLM resource management programs
helps us better comprehend how BLM Carlsbad Field Office is achieving its mission of multiple-
use and sustained yield of resources and resource uses.
3

1.1. Study Area
The BLM Carlsbad Field Office (CFO) in New Mexico was selected as the study region
(Figure 1) because the planning area manages multiple resources and resource uses with
contrasting management goals and objectives. The primary resource uses in BLM-administered
lands of the planning area are oil and gas extraction, potash mining, caliche mining, livestock
grazing, and off-highway vehicle (OHV) recreation. The combination of these resource uses
modifies the landscape in a variety of different ways, which warrants a thorough investigation of
how the cumulative effects affect the ecological integrity and landscape health of the land. The
CFO also provides extensive field datasets that are parameterized as indicators and metrics for
the Landscape Integrity Index.

Figure 1. Map of BLM Carlsbad Field Office Planning Area
4

1.2. Motivation
1.2.1. A Comprehensive, Standardized, and Transparent GIS Approach
With the challenge of defining the geographic (spatial) and time (temporal) boundaries of
the cumulative effects, there is a need to establish a comprehensive, standardized, and
transparent Geography Information System (GIS) approach. Comprehensive means including
relevant and scientifically sound data and analysis, standardized means using consistent and
repeatable measures, and transparent means providing clear and documented workflows. The
GIS approach can help the subject area experts and decision makers understand the ecological
integrity of BLM managed lands at a landscape level (McGarigal and Marks 1995) and provide a
framework for continual awareness, communication, and coordination of effective resource
management (Atkinson and Canter 2011). This research utilizes standalone python scripts and R
Markdown that can be shared with other BLM offices. With a shareable GIS workflow, other
field offices can customize it according to their management and data needs and share their
findings with other offices. The GIS workflow will promote standardized CEA measures and
collaborative management across administrative boundaries.
1.2.2. Evaluation of Ecological Integrity in CEA
The ability to quantify and evaluate ecological integrity helps establish a holistic
framework for measuring the effects of the resource programs and communicating the progress
of the multiple-use and sustained yield mission to managers and stakeholders (Wurtzebach and
Schultz 2016). Identifying indicators to measure elements of ecological integrity is key to
evaluating the effects of the past, present, and future resource management programs. Using
indicators and indices to evaluate cumulative effects of multiple actions has started to gain
traction (Canter and Atkinson 2011), and land-management agencies are using indicators to
5

evaluate ecological integrity (Carter et al. 2019). That said, no formalized LII has been
developed to evaluate cumulative effects of BLM resource management programs on ecological
integrity. This research will bridge the two frameworks – cumulative effects analysis and
evaluation of ecological integrity – together through the creation of LII. Three components
incorporating ecological integrity, management, and landscape will form the LII: (1) ecological
indicators of ecosystem composition, structure, and function, (2) resource- metrics and stressor-
based metrics, and (3) landscape metrics.
1.2.3. Landscape Metrics
Incorporating landscape metrics in the LII can help us capture the complex spatial
patterns and interactions influenced by the multiple-use resource management programs over
time (McGarigal and Marks 1994). The landscape metrics represent landscape structure, one of
the three characteristics of the landscape, and define the spatial relationships between diverse
ecosystems. Landscape structure is portrayed by the spatial pattern characterized by landscape
composition (the variety and abundance of patch types within a landscape) and configuration
(spatial characteristics of patches within the landscape), with patches being the basic elements or
units that make up a landscape (McGarigal and Marks 1994, Uuemaa et al. 2009). Being able to
quantify landscape structure through ecological similarity and connectedness with landscape
metrics allow us to incorporate the interactions between ecological processes and landscape
dynamics (McGarigal and Marks 1994).
Moreover, landscape metrics can also help us examine landscape fragmentation and
diversity as effects from the resource management programs. Landscape metrics quantify
landscape structure at the patch-, class-, and landscape-levels, meaning measurements are
performed for each individual patch, all patches belonging to the same land cover class type, and
6

all patches in the landscape (regardless of land cover type), respectively (Frazier 2019). Many of
the class-level metrics can provide information on landscape fragmentation while the landscape-
level metrics can provide information on landscape diversity. Landscape fragmentation occurs
when the resource use activities subdivide the ecosystems into smaller and isolated fragments
that can reduce biodiversity (McGarigal and Marks 1994). This process transforms the landscape
through changes in landscape composition, structure, and function, and can negatively affect
ecological processes and habitat patches. Another aspect of this study is landscape diversity, in
which the resource use activities improve landscape health by generating more diverse
landscapes and promoting biodiversity. The inclusion of landscape metrics as part of the LII
provides a landscape perspective as to how resource use activities affect the ecological processes
spatially and temporally.
1.2.4. Multiscale Data
The framework of the Landscape Integrity Index can utilize multiscale data, which is
crucial for implementing a landscape approach in managing public lands. To understand the
effects of scale-specific ecological processes and interactions, a multiscale view of landscape is
necessary to apply multiscale approaches for ecosystem modeling. In addition to spatial scale
(i.e. grain: resolution of the data; and extent: size of the landscape), temporal scale is also an
important consideration in assessing changes in the landscape over time. Furthermore, field data
have localized details that can help management at the field office level, while national data
provides a standardized assessment that can help management at the landscape level. The BLM
Assessment, Inventory, and Monitoring (AIM) process is a national monitoring effort that
integrates local- and broad-scale data collection and provides the status and condition of natural
resources (Carter et al. 2017). Since LII can be developed using multiscale data, if some of the
7

field offices do not have the data or no AIM data exists within the field office boundary, the LII
can still be developed using existing field or national datasets, whichever is available.
1.2.5. Court Challenges
Given the complex and unanticipated nature of cumulative effects, aggregate impacts are
not consistently translated into clear and transparent guidance for CEA professionals to apply in
practice (Foley et al. 2017). Court challenges relating to CEA against federal agencies have been
raised due to the lack of cumulative impact analysis, lack of past, present, or reasonably
foreseeable future actions, lack of data and/or credible justification for selection of data, and
unsubstantiated assertions that there are no cumulative impacts from the projects (Smith 2006).
Federal agencies have lost many of those challenges because of the difficulties in conducting a
comprehensive, systematic, and transparent cumulative effects analysis. Smith (2006) found that
the Bureau of Land Management had lost all three of their cases from the Ninth Circuit Court of
Appeals from 1995 to 2004. In Klamath-Siskiyou Wildlands vs. BLM (2004), for example, the
court ruled that the CEA was inadequate for the two timber sales in the Cascade Mountains of
southern Oregon because there was not enough analysis of other timber sales in the same
watershed, CEA lacked data and justification, and analysis cannot be tied to a RMP with no site-
specific analysis nor to a non-NEPA document. This case demonstrates the importance of data
and the need for a reliable rationale for the selection of data. The standardized LII framework
and transparent GIS workflow from this research can be used by land managers and CEA
practitioners to improve the CEA process.
1.3. Research Purpose, Goals, and Objectives
The purpose of this thesis was to develop a Landscape Integrity Index (LII) with
ecological integrity indicators, resource- and stressor-based metrics, and landscape metrics
8

(Figure 2) to evaluate the cumulative effects of BLM resource management programs. The
research aims to improve the Cumulative Effects Analysis in the Resource Management Plans by
providing a shareable GIS workflow that is comprehensive, systematic, and transparent. The case
study is demonstrated in the BLM Carlsbad Field Office, New Mexico by analyzing just two of
the resource management programs, Vegetative Communities and Minerals – Leasables – Oil
and Gas. Both of these programs have various management plans and actions that are cast at
different spatial and temporal scales within the CFO planning area.

Figure 2. Landscape Integrity Index (LII) Framework
The goals of this research project were as follows:
(1) To create a comprehensive and shareable GIS model that evaluates the cumulative
impacts of several programs in the BLM Carlsbad Field Office, New Mexico.
(2) To incorporate ecological integrity indicators, resource- and stressor-based metrics,
and landscape metrics to create a Landscape Integrity Index (LII).
(3) To assess areas of high and low landscape integrity in BLM Carlsbad Field Office.
In order to accomplish these research goals, the objectives of this research were listed as follows:
9

(1) Research historical and current methods for cumulative effect analysis in the BLM to
identify the components needed in the analysis.
(2) Apply field data in the LII model as indicators and metrics according to the resource
management plan in BLM Carlsbad Field Office.
(3) Conduct the moving window analysis to depict the Landscape Integrity Index and use the
LII values to assess landscape integrity.
1.4. Research Questions
The research questions that drove this study were as follows.
(1) How can spatial data and standardized measures of ecological integrity and landscape
health be used to evaluate the cumulative impacts of Vegetative Communities and
Minerals – Oil and Gas?
(2) What elements of ecological integrity and landscape health should be included in the
BLM’s cumulative effects analysis to provide a more comprehensive, standardized, and
transparent assessment?
(3) How does this assessment reveal the cumulative impacts the BLM resource management
programs (specifically Vegetative Communities and Minerals – Oil and Gas) have on the
public lands in BLM Carlsbad Field Office, New Mexico?
(4) Where are the areas of low and high landscape integrity in the BLM Carlsbad Field
Office?
1.5. Thesis Structure
The remainder of this thesis consists of four chapters. Chapter 2 explores the BLM’s
mission and history of cumulative effects analyses, identifies the challenges in conducting a
cumulative effects analysis, and defines the indicators and metrics for the Landscape Integrity
10

Index (LII). Chapter 3 presents the study area, data sources and selection of indicators and
metrics, the three-stage workflow, and procedures of designing and developing the LII model.
Chapter 4 describes the results and Chapter 5 discusses the landscape integrity of CFO Planning
Area, applications of LII, research limitations, and future research opportunities.
11

Chapter 2 Related Work
The presence of the BLM can be felt in the majestic mountains towering in wilderness
areas, sheep and cattle grazing in the distance, the busy hum of drilling machines extracting oil
and gas, and people creating memorable experiences on public lands. Protecting these multiple
natural, cultural, and historic resources is an incredible and challenging undertaking for the
BLM. To appropriately manage the multiple and often competing resource management
programs, the BLM needs to understand the impacts of all of these programs by conducting the
cumulative effects analysis (CEA) during the National Environmental Policy Act (NEPA)
process. The hurdles of conducting the cumulative effects are also multifaceted, as the intricacies
of identifying spatial and temporal scale and analyzing indirect and future effects pose quite a
dilemma for CEA practitioners and land managers.
This research addresses the challenges in conducting the CEA by examining how the
resource management programs affect the landscape health and the ecosystem through the
creation of a Landscape Integrity Index. The different components in the LII framework –
ecological integrity indicators, resource- and stressor-based metrics, and landscape metrics –
provide a systematic and transparent way for evaluating the cumulative effects of resource
management programs.
2.1. BLM’s Mission
As mandated by the Federal Land Policy and Management Act of 1976 (FLPMA), the
BLM manages the public land under the principles of multiple-use and sustained yield, while
protecting the scientific, scenic, historical, ecological, environmental, air and atmospheric, water
resource, and archeological values of the lands (U.S. Department of the Interior 2001). The BLM
supports a wide range of resource management programs such as energy development, livestock
12

grazing, timber harvesting, conservation stewardship, and recreation, utilizing a multiple-use
approach and striving for a long-term sustainable management of the public lands (Carter et al.
2019). Some of these programs may disrupt landscape patterns and reduce ecologic integrity and
landscape health, while others may improve them or mitigate the negative impacts. For instance,
conventional energy development such as coal mines and oil and gas drilling modify the
structure of surface landscapes and disrupt vital ecological processes, resulting in increased
habitat fragmentation and decreased biodiversity with long-term consequences (Copeland et al.
2009, Allred et al. 2015, Wu et al. 2019). On the other hand, habitat restoration activities such as
vegetative treatments can improve patch size and connectivity of habitat, ultimately improving
the landscape health for that region (Carter et al. 2017). As such, incorporating measurements of
ecological integrity and landscape health are essential in management decisions to help sustain
and protect resources and resource uses despite competing resource management goals.
2.1.1. Landscape Approach
Effective management of the diverse range of resource management programs throughout
the various BLM offices involves collaboration with government agencies and organizations and
comprehension of the program effects by managers and stakeholders (Carter et al. 2017). This is
currently being addressed by implementing a landscape approach to resource management, in
which the BLM engages with diverse stakeholders, considers the resource values and tradeoffs in
operating different resource management programs, and incorporates multiscale and broad-scale
spatial and temporal perspectives (Carter et al. 2017). Included in the landscape approach effort
is the multiscale natural resource monitoring and assessment information, which aligns with the
BLM Assessment, Inventory, and Monitoring (AIM) process to inform BLM resource planning
and management decisions (Carter et al. 2017). The challenge lies in meeting the objectives of
13

multiple-use resource management programs while protecting the ecological integrity and health
of public lands.
2.2. Cumulative Effects Analysis
2.2.1. History of Cumulative Effects Analysis
The National Environmental Policy Act (NEPA) of 1969 requires the BLM to conduct
cumulative effects analysis of a proposed action to assess the incremental impact of past, present,
and reasonably foreseeable future actions on the environment (Bureau of Land Management
2008). But throughout the 15 years following the inception of NEPA, many agencies have failed
to include CEA or submit a well-written CEA in the NEPA documents, leading to an increase in
the court cases challenging cumulative effects analyses (Smith 2006). The main problem arose
from the lack of clear guidelines, scopes, and proper procedures for preparing a cumulative effect
analysis (Canter and Kamath 1995). In 1997, the Council of Environmental Quality published
the “Considering Cumulative Effects Under the National Environmental Policy Act”, which is a
handbook that provides a framework and process of analyzing cumulative effects (U.S. Council
of Environmental Quality 1997). However, cumulative effects analysis still remains confusing
for NEPA practitioners even with the publication of CEQ’s handbook (Smith 2006).
2.2.2. Defining Cumulative Effects
Analyzing the cumulative effects of BLM program actions is complex and challenging
because it is difficult to keep track of cumulative effects when they can be produced and interact
in multiple ways (Foley et al. 2017). For example, a single activity can repeatedly produce a
single stressor or multiple stressors, and multiple activities can produce a single stressor or
multiple stressors. A stressor is the environmental and biotic factor created from human activities
that causes stress to an ecosystem. Response to the stressor can be affected by additional
14

stressors. For instance, a species can respond differently to invasive species under different
nutrient conditions (Crain et al. 2008). In cases where stressor A reduces the response by ‘a’ and
stressor B by ‘b’, the cumulative effects of multiple stressors can then interact in multiple ways:
additively (cumulative effects = a + b), synergistically (cumulative effects < a + b), or
antagonistically (cumulative effects > a + b). Figure 3a shows how cumulative effects are
produced by multiple stressors from multiple activities and can interact with each other (dashed
lines between effects arrows). Figure 3b shows that the impact is characterized by activity,
stressor, and/or by ecological components (dashed lines). The terms “effects” and “impacts” are
synonymous according to the CEQ regulations. Effects are changes that result from action(s) and
can be ecological, aesthetic, historic, cultural, economic, social, or health (Bureau of Land
Management 2008). Effects can also be beneficial or detrimental, and short-term or long-term.
To evaluate the cumulative effects of BLM resource management programs, this research
considers and scores each indicator and metric, then combines these scores into a Landscape
Integrity Index to produce a value for the cumulative effects.

Figure 3. (a) Relationships between activities (from resource management programs), stressors,
and ecological components; (b) relationships between impact characterization and activity,
stressor, and/or ecological components. Source: Foley et al. (2017)
15

2.2.3. Evaluating Cumulative Effects
In order to improve CEA, definition of impact and relationship between activities,
stressors, and ecosystem effects need to be consistently established with the best available
science (Foley et al. 2017). Foley et al. (2017) evaluated how CEA practitioners conduct CEA,
specifically examining key gaps in and relationships between impacts, baseline, scale, and
significance in a comparative case analyses in California, Canada, Australia, and New Zealand.
Baseline condition is the condition of the ecosystem prior to human impact and is used to
compare ecosystem effects with and without the resource management programs. Some of their
recommendations to improve CEA include mapping overlapping and potentially interactive
effects to assess impacts; increasing access to data and details of past, present, and future
projects across jurisdictional boundaries to define baseline; improving understanding of
threshold dynamics and feedback loops and incorporating chronic impacts that act over long
temporal scales to define spatial and temporal scale; and developing ecological indicators that
signify wide-ranging ecosystem change to determine significance.
In my study, impacts, spatial and temporal scale, and significance are incorporated in the
data and methodology of developing the LII. Impacts consider both the activity (i.e. oil and gas
wells drilling) and impact type (i.e. habitat disturbance). Spatial and temporal scales are
determined by the jurisdiction boundary of the BLM Carlsbad Field Office, distance of influence
from BLM staffs and research literature, fiscal year of operation, and availability of data. These
scales are further addressed in the methodology (Section 3.3.2). Significance is defined as
“effects of sufficient context and intensity that an environmental impact statement is required” in
the NEPA Handbook. In other words, the action is significant in context to society as a whole,
the affected region, the affected interests, and locality, and the severity of effect is significant.
Areas with LII value higher than 0.8 are be referred to as significant areas of high landscape
16

integrity, and areas with LII value lower than 0.3 are referred to as significant areas of low
landscape integrity.
Carter et al. (2019) proposed a method of using indicators that consider the ecological
health (structure, composition, and function) and management objectives and policies (resource-
and stressor-based) to evaluate ecological integrity. This method sets the foundation for selecting
the ecological and management indicators for my research. In addition to those indicators, my
research also comprises landscape metrics to account for spatial patterns and processes to create
a Landscape Integrity Index.
2.3. Challenges to Cumulative Effects Analysis
2.3.1. Spatial and Temporal Scale
The implementation of GIS in cumulative effect analysis is largely focused on
establishing spatial and temporal boundaries and identifying vulnerable resources and areas
where the greatest effects occur (Atkinson and Canter 2011). Generally, the natural boundaries
of the resource affected defines the geographic scope, not the jurisdictional boundaries (Bureau
of Land Management 2008). Some challenges with the CEA involving spatial scale include
having too small of a geographic area for the analysis (Smith 2006) or using jurisdiction as the
spatial extent (Foley et al. 2017), which has resulted in missing potential important contributing
factors that affect ecological components. Moreover, determining the geographic scope of the
CEA can be difficult when there are multiple land use activities across the landscape.
The timeframe for each cumulative effect should be established by defining the long- and
short-term effects and incorporating the duration of the effects anticipated (Bureau of Land
Management 2008). However there are several challenges in selecting the appropriate temporal
scale. One of the issues from past CEA is having too short of a timeframe (Smith 2006). Other
17

issues include using the operational period of the project or the duration of the action (Foley et
al. 2017), limited availability of long-term data, or a temporal lag between the action and its
effect until triggered by rare events such as extreme weather (Harvey and Railsback, 2007). The
LII model in my research is also confined by the fiscal year of the project and the available data,
but future attempts in advancing the LII model should include additional background research on
the duration of the effects and how long the effects affect ecological components.
2.3.2. Indirect and Future Effects
There is also a gap in identifying indirect and cumulative effects from multiple past,
present, and future actions. Indirect effects are effects caused by actions that occur later in time
or further in distance that can induce changes in the pattern of land use and other ecological
processes (Bureau of Land Management 2008). Interactions among the past, present, and future
actions include additive (sum of the effects make up the cumulative effect), countervailing
(effects of some actions balance the effects of other actions), and synergistic (the total effect is
greater than sum of the individual effects) (Bureau of Land Management 2008). The GIS
workflow in this research project considers these indirect effects and interactions in the past,
present, and future by examining the spatial and temporal changes in the landscape structure and
how these changes are reflected in the spatial patterns. While it may be difficult to distinguish
between different interactions, determining the impact scores and distance of influence for the
resource management activities and the stressors will help identify the magnitude and the areal
extent of the effects. For example, identifying that oil and gas wells have an impact score of 0.2
(from previous landscape models) and calculating its Euclidean Distance with distance decay
function can help quantify the amount and distance of the impact have on the environment.
18

2.4. Ecological Integrity and Its Indicators
The scientific community and land management agencies have an increased awareness of
the concept of ecological integrity and the need to assess it as a way to manage natural systems
(Hobbs et al. 2010, Carter et al. 2019). However, most ecological indicators, including
environmental indices, and habitat suitability models have been identified for assessing the
ecological integrity of aquatic habitats, rather than terrestrial habitats (Canter and Atkinson 2011,
Carter et al. 2019). By contrast, this research considers a variety of terrestrial ecosystems by
incorporating landscape metrics and patch analysis, which can lead us to a better understanding
of the ecological integrity of different landscapes and how to better manage resource
management programs.
2.4.1. Composition, Structure, and Function
Quantifying ecological integrity and landscape health is one way to evaluate the
cumulative effects of the BLM resource management programs. The characterization of
ecological integrity at a landscape-level is described by the elements of composition, structure,
and function (Andreasen et al. 2001, Dale and Beyeler 2001), in which the health level of an
ecological system is determined by its endurance and recovery dynamics against environmental
processes or human disturbances (Parrish et al. 2003). Composition emphasizes the biological
elements that influence ecosystem processes, such as focal or indicator species, species richness
or evenness, or richness of patch size (Andreasen et al. 2001, Dale and Beyeler 2001). Structure
comprises of landscape-level elements such as physical features, habitat fragmentation, or
landscape connectivity. Function incorporates biotic and abiotic processes and interactions such
as productivity, predation, weather, or disturbance. The inclusion of these components of
19

ecological integrity in a cumulative effects analysis can better help identify the incremental
effects of BLM resource management programs have on the managed lands.
2.4.2. BLM Management Indicator Species
To identify the appropriate ecosystems and weights of ecological integrity variables for
the LII model, BLM management indicator species and their habitats were examined to
determine the general habitats (i.e. vegetation areas) and associated habitat requirements. BLM
New Mexico manages Bureau Sensitive Species (i.e. at-risk native species) and their habitats in
BLM lands by planning and implementing conservation actions to prevent species listing and
eventually remove them from the sensitive species list (Bureau of Land Management 2019).
Three amphibians, two arthropods, twenty birds, five mammals, and two reptiles species were
selected from the 2018 BLM NM Sensitive Species list, with existence verified in Carlsbad Field
Office (Appendix A). Most of these species are either Endangered, Threatened, or Species of
Greatest Conservation Need (SGCN) under the NM Status; or Watch (species of concern with
the potential to become problematic), Watch New, BLM Sensitive (BLM determined priority
species), or BLM Sensitive New under BLM Status. These species act as the BLM management
indicator species for Carlsbad Field Office, and their habitats help identify the which vegetation
area to include in the LII model (Appendix A). Additionally, given that patch size and structural
connectivity vary amongst species, Lesser Prairie Chicken was selected as the species of concern
since this species was given management actions and habitat restoration plans in the Draft BLM
CFO RMP. The habitat requirements of Lesser Prairie Chicken are used to identify specific acres
and distance to be used in the LII model. Other BLM management indicator species can be
selected in future developments of LII model to represent the needs of other species.
20

2.5. Use of Landscape Metrics
Landscape metrics are emphasized in this study as a separate landscape indicator to
characterize landscape patterns, changes, fragmentation, and diversity. Although there are
several metrics for quantifying the structural components of ecological integrity in Carter et al.
(2019), these metrics are not class-level pattern metrics used for analyzing landscape
fragmentation or landscape diversity. Wang et al. (2014) considered 9 out of 64 class-level
landscape pattern metrics to be robust for fragmentation measurements, including core area,
shape, proximity/isolation, contrast, and contagion/interspersion. They compared numerous
metrics including the ones available in FRAGSTATS to assess how aggregation affects pattern
metrics and how habitat abundance dependency affects metrics, which can influence the
selection of landscape metrics. FRAGSTATS is a program that quantifies landscape structure for
vector and raster images by generating a variety of landscape metrics for 3 groups of metrics
(patch, class, landscape mosaic), including area metrics, patch density, size and variability
metrics, edge metrics, shape metrics, core area metrics, diversity metrics, and contagion and
interspersion metrics. It is more advantageous to use the landscapemetrics package in R for
calculating landscape metrics since it provides a reproducible workflow, and uses the most
common metrics from FRAGSTATS and new metrics from the current literature on landscape
metrics (Hesselbarth 2019). Table 1 shows the major subject areas for landscape metrics
introduced from Wang et al. (2014), definitions, and how they provide information on landscape
fragmentation or diversity. Since the prevention of habitat fragmentation, loss of habitat, and
improving habitat diversity are key management goals at the BLM, landscape metrics that assess
habitat fragmentation and diversity are included in this study.

21

Table 1. Subject Area for Landscape Metrics, Definitions, and Information Relevant to
Landscape Fragmentation or Landscape Diversity
Subject Area for
Landscape Metrics
Definition Landscape Fragmentation /
Landscape Diversity
Shape Metrics Quantifies landscape
configuration in terms of the
complexity of patch shape.
Simple patch shape may be a result of
human-induced fragmentation.
Generally, patch shape should be
geometrically complex in natural,
unaltered landscape.
Core Area Metrics Quantifies landscape
composition and landscape
configuration in terms of the
core area of a patch.
Core area can serve as habitat area, in
which the values from the core area
metrics can show if landscape is a more
fragmented configuration of habitat and
contains less suitable habitat.
Proximity/Isolation
Metrics
Quantifies landscape
configuration through the
placement of patch types
relative to the same patch
type within a specified
distance.
Lox proximity/high isolation means
that the habitat is far away from the
same patch type, which can
characterize fragmented habitats. As
habitat diminishes and becomes
fragmented, the remaining habitat
becomes more isolated from each other
in space and time.
Contrast Metrics (see
edge metrics in 1995)
Quantifies landscape
configuration through the
degree of contrast among
patch types
High degree of contrast between
patches can indicate fragmentated
habitat with boundaries between
different patch types.
Contagion/Interspersion
Metrics
Quantifies landscape
configuration through patch
type interspersion (i.e. the
intermixing of units of
different patch types) and
patch dispersion (i.e. the
spatial distribution of a patch
type). In other words,
contagion measures the
extent to which patch types
are aggregated or clumped.
Higher values of contagion characterize
landscapes with a few large, contiguous
patches (low fragmentation), whereas
lower values generally characterize
landscapes with many small and
dispersed patches (high fragmentation).
Diversity Metrics Quantifies landscape
composition through 2
components: richness
(number of patch types
present) and evenness
(distribution of area among
different types).
Higher values from diversity metrics
suggest more number of patch types
and even area distribution among patch
types, which can indicate greater
landscape diversity.
Source: McGarigal and Marks (1995, McGarigal (2015)
22

2.6. Landscape Integrity Index
Several studies have been done to develop a landscape index of ecological integrity using
measures of human footprint (i.e. human modification in the environment), indicators of
ecological integrity, and intactness and resiliency metrics (Andreasen et al. 2001, McGarigal et
al. 2018, Walston et al. 2018), but none of them has been applied to cumulative effects analysis.
For instance, McGarigal et al. (2018) created the index of ecological integrity (IEI) by
combining anthropogenic stressor metrics representing intactness and resiliency in a weighted
linear model, in which the change in IEI over time computed the index of ecological impact.
Walston et al. (2018) developed a Landscape Integrity Index as a landscape indicator of
ecological integrity using measures of human modification on the environment (e.g. human
footprint), and indicators of ecological integrity including biodiversity (e.g. species richness) and
landscape change (e.g. vegetation departure). His paper introduced the methodology of moving
window analysis to compute LII, which will be used in my research and further explained in
Section 3.4.4.
The qualities of the ideal characteristics of a Landscape Integrity Index should include
comprehensiveness, multi-scale, naturalness, relevancy, helpfulness, integration of aquatic and
terrestrial ecology, flexibility, and measurability (Andreasen et al. 2001). These qualities go in
tandem with the qualities of scientifically sound, transparent, comprehensive methods of
conducting CEA. Using LII to evaluate the resource management programs at BLM offices will
provide a standardized and comprehensive measure of ecological integrity and landscape
patterns, which will help inform management and conservation decision making.
The use of a composite index of ecological integrity has been criticized in studies due to
the loss of information, the inability to explore individual factors that affect ecological integrity,
23

statistical problems, and the masking of variation in the direction and magnitude of effects of
individual metrics to stressors (Carter et al. 2019). These authors suggested presenting individual
metrics to managers and not combining the metrics. This study advocates the use of Landscape
Integrity Index in cumulative effects analysis because the combination of the indicators and
metrics better represents the cumulative effects on land disturbance and health.

24

Chapter 3 Methodology
The purpose of this study was to develop a Landscape Integrity Index (LII) to evaluate
the cumulative effects of BLM resource management programs, specifically Vegetative
Communities and Minerals – Leasables – Oil and Gas. This chapter first describes the study area,
then the data sources, parameters, and the selection of indicators and metrics for the LII model.
The next section applies the three-stage workflow developed by Carter et al. (2019): 1)
Understand the management context of BLM Carlsbad Field Office; 2) Design the LII model
using target resources, key stressors, spatial and temporal scales, and landscape metrics of LII
model; and 3) Implement the LII model. The procedures for conducting the data preparation,
landscape metrics in R, composite scoring system, moving window analysis, and model
validation are explained further in the chapter.
3.1. Study Area
The study area was the BLM Carlsbad Field Office (CFO) in New Mexico (Figure 1), in
which the LII model was applied. The CFO planning area is in southeastern New Mexico, within
the Eddy, Lea, and a portion of Chaves County (Bureau of Land Management 2018b). The
landscape of the planning area is predominantly desert with parts of three ecoregions including
the Chihuahua Desert, Arizona/New Mexico Mountains, and High Plains. Approximately 2.1
million surface acres of federal land out of the estimated 6.2 million acres in the planning area
will be affected by the decisions in the approved resource management plan (RMP). Neighboring
BLM field offices can use this study area as a starting template for identifying indicator species
and ecological indicators.
25

3.2. Data Sources and Selection of Indicators & Metrics
The BLM Carlsbad Field office provided the main datasets for the two resource
management programs as they are the office responsible for the Draft RMP. Table 2 summarizes
the spatial data inputs for the management actions and measures in Vegetative Communities and
Minerals – Oil and Gas. Each of the management actions and measures was categorized into an
ecological indicator, resource-based metric, stressor-based metric, or landscape metric.
Table 2. Spatial data inputs of management action and measure for Ecological Indicators and
Resource- and Stressor-based Metrics
Management Action/Measure Indicator/Metric Data/Sources
Vegetative Communities
Vegetation area
(2001, 2008, 2010, 2012, 2014)
Ecological Integrity Indicator Existing vegetation type
(LANDFIRE 2016b)
Vegetation alteration
(2001, 2008, 2010, 2012, 2014)
Ecological Integrity Indicator Vegetation departure
(LANDFIRE 2016a)
Patch size
(2001, 2008, 2010, 2012, 2014)
Ecological Integrity Indicator Existing vegetation type
(LANDFIRE 2016b)
Structural connectivity
(2001, 2008, 2010, 2012, 2014)
Ecological Integrity Indicator Existing vegetation type
(LANDFIRE 2016b)
Important plant areas
(2017)
Ecological Integrity Indicator Important plant areas of
New Mexico (EMNRD
2017)
Noxious weed treatment areas
(2001-2016)
Resource-based Metric Noxious Weed Treatment
Areas (BLM 2018b)
Vegetative Treatment
(2002-2018)
Resource-based Metric CFO VTRT Data
(BLM 2018b)
Oil and Gas Wells
Oil and gas wells
(2001-2014)
Stressor-based Metric Existing oil and gas wells
(BLM 2018b)
Applications for Permit to Drill
(APDs) (2001-2018)
Stressor-based Metric apd point
(BLM 2018b)
Flowline
(2011-2018)
Stressor-based Metric apd line
(BLM 2018b)
Pipeline
(2011-2018)
Stressor-based Metric apd line
(BLM 2018b)
Powerline
(2011-2018)
Stressor-based Metric apd line
(BLM 2018b)
Road
(2011-2018)
Stressor-based Metric apd line
(BLM 2018b)
26

Frac pond
(2009-2018)
Stressor-based Metric apd polygon
(BLM 2018b)
Well pad
(2009-2018)
Stressor-based Metric apd polygon
(BLM 2018b)
3.2.1. Ecological Integrity Indicators
As indicators of land health and ecological integrity, vegetation area, vegetation
alteration, patch size, and structural connectivity were selected as ecological integrity indicators
to quantify the compositional, structural, and function components of ecological integrity (Carter
et al. 2019). Important Plant Areas was also included as an ecological integrity indicator in this
analysis.
Existing Vegetation Type (EVT) and Vegetation Departure (VDEP) datasets (both 30 m
resolution) were from the Landscape Fire and Resource Management Planning Tools Project
(LANDFIRE). Existing Vegetation Type represented the plant community types that occurred in
the location (LANDFIRE 2016b). Vegetation area was categorized by the EVT_PHYS and
SYSTMGRPPH fields (depending on the year) from the EVT dataset, in which “Conifer”,
“Conifer-Hardwood”, “Grassland”, “Riparian”, and “Shrubland” were selected to represent the
general habitats of the BLM management indicator species and ultimately the different
ecosystems within CFO. Vegetation alteration was from the VDEP dataset, depicting the changes
in species composition, structural stage, and canopy closure between current vegetation
conditions and reference vegetation conditions (pre-EuroAmerican settlement) (LANDFIRE
2016a).
Patch size (acres) and structural connectivity (meters) were calculated for each vegetation
area. Patch size refers to the amount of vegetation area, and it is a measure of landscape
configuration (McGarigal and Marks 1995). Patch size could affect the ecological properties of a
patch via the surrounding neighborhood (e.g. edge effects). Structural connectivity is the
27

proximity of vegetation area; for instance, grassland connectivity is the distance between
individual grassland pixels. Connectivity could affect the permeability of various patch types,
movement of organisms, and ecological processes and interactions (McGarigal and Marks 1995).
Even though the Lesser Prairie Chicken is a highly mobile species, its broods have limited
mobility and their habitats need to be close to nesting habitats (Pelt et al. 2013). Ideal habitat for
the species is a mosaic of brood and nesting habitat, with a distinction between the forb and grass
cover. However, this distinction requires additional datasets that are not currently available.
Therefore, the habitat patch for Lesser Prairie Chicken is generalized and connectivity in this
study is limited to linear distance between the habitat patch. Moreover, the RMP identified a
minimum patch size of 320 acres without fragmentation from development as habitat
requirement. Other research shows that grassland is the habitat of Lesser Prairie Chicken, with a
minimum habitat patch size ranges from 4,900 ha (12,108.16 acres) to 20,236 ha (50,004.245
acres) (Spencer et al. 2016), and < 2 miles (3,218.69 m) between habitat patches for connectivity
(Pelt et al. 2013). Hence only grassland patch size and connectivity with specific range of acres
and meters were used in the analysis.
New Mexico State Forestry was the source of the Important Plant Areas of New Mexico
data. Important Plant Areas (IPA) are specific areas that support either a high diversity of
sensitive plant species or are the last remaining habitats of rare and endangered plants in New
Mexico (Natural Heritage New Mexico n.d.). This variable was included to incorporate habitats
of sensitive plant species and highlight the importance of these areas. Note that there was no
evaluation of habitat or landscape integrity for this dataset, and some of the polygons include
towns, mines, roads, and other heavily impacted areas.
28

3.2.2. Resource- and Stressor-based Metrics
Resource-based metric measures resources and resource uses, often those that are
managed by the agency (e.g. vegetation, soils, and etc.). In this study, noxious weed treatment
areas and vegetative treatment data layers were selected as resource-based metrics to assess how
these management actions affect the Vegetative Communities. The noxious weed treatment areas
are transect lines where the herbicide is applied. The vegetative treatment includes areas of
different treatment such as removal of invasive species, the use of fertilizer, pesticide, and other
treatments.
Stressor-based metrics measure anthropogenic drivers that affect the ecosystem (e.g. oil
and gas operations, grazing, etc.). Oil and gas wells, applications for permit to drill (APDs)
points, linear features of approved APD (i.e. flowline, pipeline, powerline, road), and polygon
features of approved APD (i.e. frac pond and well pond) were selected as stressor-based metrics
to assess how these management actions affect the landscape.
3.2.3. Landscape Metrics
Landscape metrics quantify the landscape structure and assess the changes in the
landscape over time (McGarigal and Marks 1995). The dataset for the landscape metrics was the
National Land Cover Database (NLCD), a 30 m resolution land cover raster dataset from the
Multi-Resolution Land Characteristics (MRLC) Consortium (Homer 2015). Appendix B lists the
NLCD land cover classes. Table 3 shows the FRAGSTATS metrics, the counterpart R-function,
definition of the metric, and interpretation of the output values from 2001, 2004, 2006, 2008,
2011, 2013, 2016, which are the years of the available NLCD data. Shape metrics, core area
metrics, contagion/interspersion metrics, and diversity metrics were selected as the final
landscape metrics in this study to analyze landscape fragmentation and diversity.
29

Proximity/isolation metrics and contrast metrics were not included because the landscapemetrics
package in R did not include functions to calculate those variables from Wang et al. (2014) (R-
function = NA in Table 3). And since some of the landscape metrics presented redundant
information insofar as they measured a similar or identical feature of landscape pattern, only a
handful (landscape metrics with a * in Table 3) were used in the development of LII. The
CAI_CV variable was used instead of CAI_SD as the measure of variability because coefficient
of variation (CV) was a relative measurement (i.e. variability expressed as a percentage of the
mean) and easier to interpret compared to standard deviation (SD), which was an absolute
measurement and interpretation was dependent on the mean (McGarigal and Marks 1995).
Landscape diversity metrics were not used in the development of LII because they were
landscape-level metrics, meaning that the output value was for the entire study area instead of for
each land cover class and would be used for comparing the landscape between different years.
The results would be discussed more in Section 4.2.
Table 3. Landscape Metrics from FRAGSTATS and R with Definition and Interpretation
(2001, 2004, 2006, 2008, 2011, 2013, 2016)
FRAGSTATS
Metrics
R-function Definition Interpretation
Shape Metrics
Perimeter-Area
Fractal
Dimension
(PAFRAC)*
lsm_c_pafrac Describes the patch complexity
of the class while being scale
independent.
1 = simple perimeter,
2 = complex shape
Core Area Metrics
Area Weighted
Mean Core
Area Index
(CAI_AM)
NA Quantifies core area for the
entire class or landscape as a
percentage of total class or
landscape area, respectively, by
weighting patches according to
their size.

0 = patch contains no
core area,
100 = patch contains
mostly core area
30

Core Area Index quantifies the
percentage of the patch that is
comprised of core area.
Coefficient of
Variation of
Core Area
Index
(CAI_CV)*
lsm_c_cai_cv Summarizes each class as the
coefficient of variation of the
core area index of all patches
belonging to the class.
Higher CV = larger
relative variation in
CAI
Standard
Deviation of
Core Area
Index
(CAI_SD)
lsm_c_cai_sd Summarizes each class as the
standard deviation of the core
area index of all patches
belonging to the class.
Higher SD = larger
absolute variation in
CAI
Coefficient of
Variation of
Core Area
(CORE_CV)*
lsm_c_core_cv Equals the coefficient of
variation of the core area of each
patch in the landscape.

Core Area is the interior area of
patches beyond some specified
edge distance or buffer width.

Higher CV = larger
relative variation of
patch core areas
Proximity/Isolation Metrics
Coefficient of
Variation of
Proximity
Index
(PROX_CV)
NA Summarizes each class as the
coefficient of variation of the
proximity index of all patches
belonging to the class.

Proximity Index considers the
size and proximity of all patches
whose edges are within a
specified search radius of the
focal patch and measures both
the degree of patch isolation and
the degree of fragmentation of
the corresponding patch type
within the specified
neighborhood of the focal patch.
Higher CV = larger
relative variation in
PROX
Contrast Metrics
Area Weighted
Mean
Euclidian
Nearest
Neighbor
Index
(ECON_AM)
NA Calculates the Euclidean Nearest
Neighbor Index by weighting
patches according to their size.

Nearest neighbor distance is
defined as the shortest straight-
line distance from a patch to the
nearest neighbor of the same
class.
Mean distance to the
nearest neighbor patch
of the same class
weighted by area
31

Total Edge
Contrast Index
(TECI)
NA Quantifies edge contrast as a
percentage of maximum
possible (landscape as a whole).

Edge contrast is the degree of
contrast a patch has compared to
its neighbor.
0 = no edge in the
landscape (entire
landscape and
landscape borders is
one patch),
100 = edge is
maximum contrast
Contagion and Interspersion Metrics
Clumpy Index
(CLUMPY)*
lsm_c_clumpy Equals the proportional
deviation of the proportion of
like adjacencies involving the
corresponding class from that
expected under a spatially
random distribution.
-1 = patch is
maximally
disaggregated,
0 = patch is distributed
randomly,
approaches 1 = patch is
maximally aggregated
Diversity Metrics
Shannon's
Diversity
Index (SHDI)
lsm_l_shdi Characterizes diversity for the
class and accounts for both the
number of classes and the
abundance of each class.

Sensitive to rare classes.
Higher = greater
number of classes;
greater diversity
Simpson's
Diversity
Index (SIDI)
lsm_l_sidi Characterizes diversity for the
class and is less sensitive to rare
class types than SHDI.

Calculates the probability that
two randomly selected cells
belong to the same class.

Responsive to dominant classes.
0 = only one patch is
present,
approaches 1 = greater
number of classes;
greater diversity
Patch Richness
(PR)
lsm_l_pr Measures the number of patch
types present; not affected by
the relative abundance of each
patch type (rare or common
patch types contribute equally to
richness) or the spatial
arrangement of patches.
Higher = more different
patch types
Patch Richness
Density (PRD)
lsm_l_prd Measures the number of patch
types per area.
Higher = more different
patch types per area
Shannon's
evenness index
(SHEI)
lsm_l_shei Calculates the ratio between
SHDI and the maximum of
SHDI.

0 = only one patch
present; no diversity,
approaching 1 =
proportion of classes is
32

An even distribution of area
among patch types results in
maximum evenness.
completely equally
distributed; greater
evenness
Simpson's
evenness index
(SIEI)
lsm_l_siei Calculates the ratio between
SIDI and the maximum of SIDI.

0 = only one patch
present,
1 = proportion of
classes is completely
equally distributed;
greater evenness
* These metrics were used in the development of LII model
Source: McGarigal and Marks (1995), McGarigal (2015), Hesselbarth (2019)
3.3. Three-Stage Workflow
My workflow was based on the three-stage method developed by Carter et al. (2019),
which assessed ecological integrity for multiple-use systems and helped inform land
management. The first stage identifies management policies and actions; the second stage
identifies target resources and key stressors to be managed, and spatial and temporal scales; and
the third stage conducts the analysis and uses the results to inform planning and management
(Figure 4). Not all steps shown in Figure 4 were adopted in my analysis, but they could be
included in future research.
33

Figure 4. 3-Stage Workflow to Assess Ecological Integrity. Source: Carter et al. (2019)
3.3.1. Understand management context
The first stage of assessing ecological integrity for multiple-use ecosystems is to
understand the management context for the assessment area (Carter et al. 2019). The
management goals and objectives in the BLM Carlsbad Field Office’s Draft Resource
Management Plan and Environmental Impact Statement (Draft RMP/EIS) provided guidance as
to the management of the CFO planning area. Table 4 showed some of the goals and objectives
in Vegetative Communities and Minerals – Leasables – Oil and Gas from the CFO Draft
RMP/EIS that were pertinent to this research with available data. By understanding the
management context for the assessment area, the Landscape Integrity Index can better
incorporate useful variables, capture the influence of management actions on landscape health,
and evaluate management effectiveness.
34

Table 4. Management Goals and Objectives of Vegetative Communities and
Minerals – Oil and Gas
Vegetative Communities
Goals • “Manage vegetation to restore the resiliency of ecosystem structure and
function, reduce fragmentation of habitat for native species, and move toward
desired plant communities.”
• “Manage public lands to prevent, eliminate, or control noxious weeds and
invasive plants.”
Objectives • “Manage public lands to prevent, eliminate, or control noxious weeds and
invasive plants.”
• “Manage for vegetation restoration, including control of undesirable and
invasive plant infestations (native and non-native species) to achieve healthy,
sustainable rangeland ecosystems that support resource values such as, but
not limited to, wildlife habitat and functional watersheds.”
• “Protect special status plant species and their habitats.”
• “Minimize or halt the spread of noxious, non-native, and invasive plant
species.”
• “Control or eliminate existing populations of noxious, non-native, and
invasive plant species. Monitor the spread of noxious, non-native, and
invasive plant species.”
• “Manage uses and treat noxious weeds such that there is no net increase in
the number of acres containing noxious weeds and reduce the number of
noxious weed species present.”
Minerals – Leasables – Oil and Gas
Goals • “Promote and support American agriculture and provide jobs and economic
development opportunities to the local community. (Executive Order 13790,
April 25, 2017).”
• “Support the national interest to promote clean and safe development of our
Nation’s vast energy resources, in a manner that does not unnecessarily
encumber energy production, constrain economic growth, and prevent job
creation. (Executive Order 13783 March 28, 2017).”
Objectives • “Allow the oil and gas industries reasonable opportunities to lease and
explore, while protecting sensitive areas and various other resources.”
• “The BLM would seek the input of industry and the public at every
opportunity to discuss changes in policy or priorities.”
• “Facilitate reasonable, economical, and environmentally sound exploration
and development of leasable minerals where compatible with resource
objectives and as consistent with Secretarial Order 3324.”
Source: Bureau of Land Management (2018b)

For example, in Vegetative Communities, the management goals and objectives specify
vegetative restoration, the prevention, elimination, and control of noxious weed and invasive
35

plants, and the protection of special status plant species and their habitats. The management
actions and field datasets (vegetative treatment and noxious weed treatment), and dataset
provided by other organization (Important Plant Areas) could attest to those management goals
and objectives. While managing for vegetation restoration is a high priority for CFO, fostering
economic opportunities from oil and gas industries is also imperative for the local community.
These two contrasting resource management programs create a dichotomy of effects that may or
may not balance each other.
3.3.2. Design assessment
The second stage is to design assessment by identifying target resources, key stressors,
spatial and temporal scales, and selecting and evaluating the metrics for the Landscape Integrity
Index. This research reviews the resource management programs in the Draft RMP and identified
the target resource as Vegetative Communities: Upland Vegetation, Noxious Weeds, and
Invasive Species, and the key stressor as Minerals – Leasable – Oil and Gas. It is important to
note that a landscape is defined as “an interacting mosaic of patches relevant to the phenomenon
under consideration (at any scale)” (McGarigal and Marks 1995), and the CEA practitioner or
land manager needs to define landscape pertinent to their management endeavor. The landscape
for my study was the BLM Carlsbad Field Office. And although patches should also be defined,
the NLCD land cover class (e.g. Deciduous Forest, Shrub/Scrub, Woody Wetlands, etc.) defined
the patches in this case.
In this study, the spatial and temporal scales were determined by the jurisdiction
boundary of the BLM Carlsbad Field Office, fiscal year of operation, and availability of data.
The spatial extent is the CFO planning area (Figure 1). Even though the spatial scale should
extend to the specific resource or ecosystem being impacted and possible ecosystem impacts
36

may exist outside of the CFO planning area boundary, funding and allocation of resources and
available data were determined by the jurisdiction boundary, and therefore the CFO planning
area boundary defined the general spatial extent. For future research, it would be ideal to obtain
field datasets from other BLM field offices and establish a spatial extent that aligned with
ecosystem processes or habitats of species of concerns to define a more natural spatial extent.
The temporal scale ranged from 2001 to 2018, depending on data availability. Some of the
datasets did not have the whole 18 years, such as the EVT and VDEP datasets (2001, 2004,
2010, 2012, and 2014), IPA data layer (2017), and NLCD data layer (2001, 2004, 2005, 2008,
2011, 2013, and 2016), which might reduce the accuracy and completeness of the LII value since
the datasets from the missing years were not accounted for. The IPA variable was the only data
layer that was used with other ecological indicators that were in different years. This might have
increased the LII value if no other ecological indicators were present at the IPA. On the other
hand, there were also occasions when there was no data in the CFO datasets (e.g. noxious weed
treatment or oil and gas wells) for specific years. If it was the case that there was no activity that
year, then the LII value would reflect the lack of activity. However, if it was the case that there
was activity that year but was not captured in the data, then the LII value would be affected by
the missing data. Moreover, past, present, and future effects were taken into consideration by the
nature of the datasets. For example, existing oil and gas wells represented the past and present
effects, while APD points represented potential past, present, and future effects.
3.3.3. Implement assessment
The third stage implements the Landscape Integrity Index through data collection and
analysis and reporting of the LII results, thereby improving the CEA process and ultimately
helping to inform planning and management actions. Since the datasets were already collected by
37

the CFO and other agencies and organizations, the main tasks were to prepare these datasets for
the analysis, perform the analysis, and validate the model. For this analysis, the LII results were
the map of the Landscape Integrity Index, a raster with LII values, and a shareable GIS model
with python scripts and R Markdown. Some ways to inform planning and management actions
would be for the CEA practitioners to look into specific areas of concerns and identify the LII
values within those areas to see if improvements can be made in those areas, if the areas should
not have future management programs, and predict the effects of potential additional
management programs.
The datasets for the ecological indicators, resource- and stressor-based metrics, and
landscape metrics were processed using a python script (LII_DataPrep.py), and this process is
explained more in Section 3.4.1. The processed NLCD data layers were the input variables to
calculate the landscape metrics using an R Markdown (LII_landscapemetrics.Rmd), and this
process is explained more in Section 3.4.2. The indicators and metrics were assigned the site
impact score or the landscape integrity values using the composite scoring system developed in
Walston and Hartmann (2018) and a python script (LII_CompositeScoringSystem.py), which is
explained in Section 3.4.3. After all of the ecological indicators, resource- and stressor-based
metrics, and landscape metrics were ranged from 0 to 1, they would be averaged into one
Landscape Integrity Index ranging from 0 (low landscape integrity) to 1 (high landscape
integrity) using the moving window analysis through a python script
(LII_MovingWindowAnalysis.py). This process is explained in Section 3.4.4. Additionally, the
LII model was validated with two comparison methods involving a linear regression model and a
Welch’s two sample t-test and using both python script and R Markdown
(LII_ModelValidation.py and LII_ModelValidation.Rmd), which is explained in Section 3.4.5.
38

3.4. Procedure
The diagram below (Figure 5) shows the general workflow, encompassing the three-stage
workflow and major steps, processes, inputs, outputs, and tools for creating the Landscape
Integrity Index. Most of the steps were performed in standalone python scripts (with ArcMap
10.6 and Python 2.7) and R Markdown (R version 3.5.3).

Figure 5. Landscape Integrity Index (LII) Workflow
39

3.4.1. Data Preparation
A series of steps was performed to prepare the data for the ecological integrity indicators,
resource- and stressor-based metrics, and landscape metrics. After going through the first two
stages of the three-stage workflow and obtaining the appropriate data, a file geodatabase for the
base variables and parameters was created. All of the raster and feature classes were then
projected to NAD 1983 UTM Zone 13N and clipped to the CFO boundary. A year text field
(Year) was added and the year information from the Last Activity field was extracted for the oil
and gas wells variable. Subsequently, raster and feature classes were extracted and selected using
different categories (i.e. vegetation area, APD lines, and APD polygons) and the appropriate
year.
For ecological integrity indicators, the patch size and structural connectivity variables
required additional steps to produce. The patch size variables for each ecosystem were created by
the Region Group tool with an eight-cell neighborhood rule. This tool identifies the patch where
each cell belongs to using the immediate surrounding cells (the eight-cell neighborhood rule
includes diagonals) (Esri 2019b). The Lookup tool was used to create a new patch size raster
using the COUNT field, which would show how many pixels were in each group. To get the
number of acres of each patch, Map Algebra (e.g. Raster Calculator tool) was used to multiply
the new raster with 0.222395 (0.222395 acres = 900 m², for 30 m pixels). The structural
connectivity variable for each ecosystem was generated by the Euclidean Distance tool, which
calculated the distance between individual ecosystem pixels (Esri 2019a).
3.4.2. Landscape Metrics in R
The landscapemetrics package in R produced the shape metrics, core area metrics,
contagion/interspersion metrics, and diversity metrics to assess landscape fragmentation and
40

diversity. The first step was setting the working directory and installing and loading the
necessary packages such as raster, rgdal, sp, kableExtra, knitr, and landscapemetrics. The inputs
were the NLCD TIFFs that were prepared in the above section. The lsm_c_pafrac function
calculated the Perimeter-Area Fractal Dimension (PAFRAC) for shape metrics. The
lsm_c_cai_cv function calculated the Coefficient of Variation of Core Area Index (CAI_CV); the
lsm_c_cai_sd function calculated the Standard Deviation of Core Area Index (CAI_SD); and the
lsm_c_core_cv function calculated the Coefficient of Variation of Core Area (CORE_CV) for
core area metrics. The lsm_c_clumpy function calculated the Clumpy Index (CLUMPY) for the
contagion/interspersion metrics. The lsm_l_shdi function calculated the Shannon’s Diversity
Index (SHDI); the lsm_l_sidi function calculated the Simpson’s Diversity Index (SIDI); the
lsm_l_pr function calculated the Patch Richness (PR); the lsm_l_prd function calculated the
Patch Richness Density (PRD); the lsm_l_shei function calculated the Shannon’s Evenness Index
(SHEI); and the lsm_l_siei function calculated the Simpson’s Evenness Index (SIEI). Additional
parameters were specified in the LII_landscapemetrics.Rmd, and the results could be viewed in
Appendix B.
3.4.3. Composite Scoring System
To model the effects each indicator and metric had on the landscape, a modified
composite scoring system from Walston and Hartmann (2018) was implemented in this analysis.
The ecological integrity indicators were characterized by only the site impact scores. The
resource- and stressor-based metrics were characterized by the modeling approach and
parameters of site impact score, distance of influence, and distance decay function, which were
adopted from previous landscape modeling efforts (Theobald 2013, Hak and Comer 2017,
Walston and Hartmann 2018). The site impact score represents the impact of the landscape
41

condition, or the ecological stress caused by the management action, and ranges from 0 (greater
site impact) to 1 (lower site impact). This definition differed slightly from Walston and
Hartmann (2018)’s definition of site impact score, which was “the assumed intensity of the
human land use.” Distance of influence is the distance at which the management action
presumed to affect ecological integrity, since habitat quality and wildlife use generally declines
with proximity to human activities. Distance decay function reveals the relationship between
ecological impact and distance from the management action, in which logistic function was used
as the distance decay function. On the other hand, the landscape metrics were not assigned any
impact score; they were only normalized from 0 (low landscape integrity) to 1 (high landscape
integrity) according to the calculated values from the landscapemetrics package.
For the ecological integrity indicators, a file geodatabase was created, and the
environmental extent was set to the CFO boundary so that the raster cells would cover the entire
CFO boundary for all variables. A site impact score double field (IP) was added and assigned a
site impact score of 1 to the IPA variable, indicating that the IPA were areas of high ecological
integrity. Since the IPA dataset was in vector format, it was converted to 30 m resolution raster
using the Feature to Raster tool. An important step was to use the IsNull tool in combination with
the Con tool to ensure that cells with Null, NoData, and area with low habitat suitability values
were set to -10 (Table 5). This step warranted that those cells would be accounted for when used
in Map Algebra or Raster Calculator tool, but the negative value would not be in the final
calculation for the LII value. Next, the Reclassify tool was used to reclassify the values of the
ecosystems to site impact score of 1 and “NoData” to -10, indicating that the vegetation areas
were areas of high ecological integrity. The Con tool was used to reassign values of -10 to
grassland patch sizes < 320 acres, 0.75 to grassland patch sizes < 12,108.16 acres, 0.95 to
42

grassland patch sizes < 50,004.245 acres, and 1 to grassland patch sizes ≥ 50,004.245 acres. In
addition, the Con tool was used to reassign value of 1 to grassland connectivity ≤ 3,218.69 m,
and -10 to grassland connectivity > 3,218.69 m. Since the vegetation alteration data layer ranged
from 0 to 100, it was inversely normalized to the range of 0 (high vegetation change from
reference vegetation condition) and 1 (little vegetation change from reference vegetation
condition) with the following equation:
Equation 1.
𝑀𝑎𝑥𝑖𝑚𝑢𝑚 – 𝑋 𝑀𝑎𝑥𝑖𝑚𝑢𝑚 − 𝑀𝑖𝑛𝑖𝑚𝑢𝑚
Raster with cell values < 0 were set to Null using the SetNull tool. The Cell Statistics tool was
used to calculate the minimum site impact score out of all of the ecological indicators for the
appropriate years.
Table 5. Site Impact Scores of Ecological Integrity Indicators

Management Action or Measure Site Impact Score NoData/Null/Area with low
habitat suitability
IPA 1 -10
Conifer 1 -10
Conifer-Hardwood 1 -10
Grassland 1 -10
Riparian 1 -10
Shrubland 1 -10
Grassland Patch Size 0.75, 0.95, 1 -10
Grassland Structural Connectivity 1 -10
Vegetation Alteration 0 - 1 NA

For resource- and stressor-based metrics, a file geodatabase was created for each
management program, and the environment extent was set to the CFO boundary. A site impact
score double field (IP) was added; and site impact score of 0.7 was assigned for the noxious
weed treatment and vegetative treatment variables, 0.2 for the oil and gas wells, APD point,
43

flowline, pipeline, frac pond, and well pad variables, 0.6 for powerline variable, and 0.75 for
road variable (Table 6).
Table 6. Site Impact Scores of Resource- and Stressor-based Metrics

Management Action or
Measure
Site Impact
Score
Null Distance of
Influence (m)
Distance Decay
Function
Resource-based Metrics
Noxious weed treatment areas 0.7 -10 500 Logistic
Vegetative Treatment 0.7 -10 500 Logistic
Stressor-based Metrics
Oil and gas wells 0.2 -10 1000 Logistic
Applications for Permit to
Drill (APDs) points
0.2 -10 1000 Logistic
Flowline 0.2 -10 1000 Logistic
Pipeline 0.2 -10 1000 Logistic
Powerline 0.6 -10 200 Logistic
Road 0.75 -10 500 Logistic
Frac pond 0.2 -10 1000 Logistic
Well pad 0.2 -10 1000 Logistic
Source: Modified from Walston and Hartmann (2018)
The site impact scores and distance of influence for resource- and stressor-based metrics were
adopted from Walston and Hartmann (2018). Noxious weed and vegetative treatment utilized the
site impact score from the “Low agriculture and invasive (ruderal forest, recently burned,
recently logged, etc.)” field, and road utilized the site impact score from the “Primitive roads
(e.g. dirt roads and trails) field from Walston and Hartmann (2018). All the resource- and
stressor-based metrics datasets were in vector format, and they were converted to 30 m
resolution raster using the Feature to Raster tool. The IsNull tool in combination with the Con
tool were used to ensure that cells with Null were set to -10. The Euclidean Distance tool was
used to calculate the Euclidean distance (meters) for each resource- and stressor-based metrics,
in which each cell value represented the distance to the closest objects of interest (in this case,
management action) (Esri 2019a). A maximum distance of 4,000 m was used in the Euclidean
Distance tool to represent the distance of influence human activities have on wildlife (Walston
44

and Hartmann 2018). The distance decay function, more specifically the Logistic 10 tool, was
applied on the Euclidean Distance output. Logistic function was selected instead of using both
linear and logistic functions. This was because most of the metrics had higher site impacts and
the difference between logistic and linear functions were minimal. The IsNull tool in
combination with the Con tool were used again to ensure that the Null in the outputs were set to -
10. The outputs from the previous step were normalized using the following equation:
Equation 2.
𝑋 − 𝑀𝑖𝑛𝑖𝑚𝑢𝑚 𝑀𝑎𝑥𝑖𝑚𝑢𝑚 − 𝑀𝑖𝑛𝑖𝑚𝑢 𝑚
The normalized raster was multiplied with the raster that had the site impact score as the cell
value to incorporate the effects from the Euclidean Distance and distance decay function have
onto the site impact scores. If there were areas that the two raster data layers did not overlap,
then the output raster cell would contain value from either the normalized raster or the raster that
had the site impact score. The IsNull tool in combination with the Con tool were used again to
ensure that the Null in the outputs were set to -10. Raster with cell values < 0 or = 100 were set
to Null using the SetNull tool to exclude Null (-10) and Null overlaps (-10 times -10 = 100). The
Cell Statistics tool was used to calculate the minimum site impact score out of all of the
resource- and stressor-based metrics for the appropriate years.
For landscape metrics, a file geodatabase was created, and the environment extent was set
to the CFO boundary. Reclassify tool was used to reclassify the NLCD values to landscape
metric values that were calculated from the landscapemetrics package in R (see Appendix B for
a complete list of the values). The raster with landscape metric values were then normalized to a
range of 0 (high landscape fragmentation) to 1 (low landscape fragmentation) with the range of
the landscape metrics or the minimum and maximum landscape metric values if the metric did
not have a range of output. For example, PAFRAC was normalized with the range of 1 to 2
45

(Equation 3), while CAI_CV was normalized with its minimum and maximum values (Equation
4):
Equation 3.
(𝑃𝐴𝐹𝑅𝐴𝐶 − 1)
2 − 1
Equation 4.
𝐶𝐴𝐼 _𝐶𝑉 − 𝑀𝑖𝑛𝑖𝑚𝑢𝑚 𝑀𝑎𝑥𝑖𝑚𝑢𝑚 − 𝑀𝑖𝑛𝑖𝑚𝑢𝑚
The Cell Statistics tool was used to calculate the minimum landscape integrity value out of all of
the landscape metrics for the appropriate years.
3.4.4. Moving Window Analysis
The LII was computed by calculating the average of all overlapping 30 m pixel values in
the raster models of ecological integrity indicators, resource- and stressor-based metrics, and
landscape metrics within 1 km moving windows using the Focal Statistics tool (Walston and
Hartmann 2018). To prepare for the moving window analysis step, a file geodatabase was
created, and the environment extent was set to the CFO boundary. The Cell Statistics tool was
used to overlay the raster models of ecological integrity indicators, resource- and stressor-based
metrics, and landscape metrics and calculate the mean of the site impact scores and landscape
metric values for each year. The Cell Statistics tool was used once more to overlay the raster
models for all of the years into one raster with mean as the statistics type. Circle was selected as
the neighborhood, and 1,000 map units (meters), mean, and ignore NoData value were selected
as the parameters for the Focal Statistics Tool. Lastly, the output Landscape Integrity Index was
clipped to the CFO boundary.
3.4.5. Model Validation
The model was validated by (1) comparing the LII values with the values from the
Landscape Condition Map (LCM) developed by Hak and Comer (2017) using a linear regression
model, and (2) comparing the LII values in protected areas to the LII values in multiple-use areas
46

using the Protected Areas Database of the United States (PAD-US) data from U.S. Geology
Survey (USGS) Gap Analysis Project (GAP) and Welch’s two sample t-test. The linear
regression would show if there was a relationship between Landscape Integrity Index and
Landscape Condition Map. And the Welch’s two sample t-test would indicate if the mean LII
value would be different between protected areas and multiple-use Areas. Both python scripts
and R Markdown were used during the model validation process. To prepare the datasets for the
model validation, a file geodatabase was created, and the environment extent was set to the CFO
boundary. All of the raster and feature classes were then projected to NAD 1983 UTM Zone 13N
and clipped to the CFO boundary. The LCM raster was manually projected due to the large size.
I then selected the areas that were protected (GAP Status = 1 or 2) and multiple-use (GAP Status
= 1 or 3) in the PADUS data and dissolved those areas. I had initially included unprotected areas
(Gap Status = 4), but there was no unprotected areas within CFO boundary, and thus it was not
included in the final model validation.
For the first model validation process, 100 random points were created within the CFO
boundary using the Create Random Points tool, then the LII values and LCM values were
extracted and recorded in those points using Extract Values to Points tool. Then the LII values
and LCM values were compared in R using a linear regression to model the relationship between
LII and LCM. In R, the working directory was set and the necessary packages such as readxl,
ggplot2, dplyr, tidyr, magittr, gridExtra, e1071, kableExtra, and knitr were installed and loaded.
The inputs were the tables of LII values and LCM values (see Appendix C). The LCM value was
the independent/predictor variable (x), and the LII value was the dependent/response variable (y).
A scatter plot was plotted to visualize the relationship between LII and LCM using the
scatter.smooth function. A density plot was used to check if the response variable was close to
47

normal using the density function. The correlation between LII value and LCM value was
calculated using the cor function. Then the linear regression model was built using the lm
function.
For the second model validation process, 100 random points was created within the
protected areas and another 100 random points was created within the multiple-use areas using
the Create Random Points tool, then the LII values were extracted and recorded in those points
using Extract Values to Points tool. The LII values in protected areas were compared to the LII
values in multiple-use areas using Welch’s two sample t-test to test the hypothesis that two
different areas have equal means. In R, the working directory was set, and the packages were
already installed and loaded from model validation process 1. Box plots were plotted to identify
any outliers within the two groups of LII values. The Welch’s two sample t-test was performed
using the t.test function.

48

Chapter 4 Results
This chapter presents the key findings of the Landscape Integrity Index model and the two model
validation processes. The results of the LII model indicate that the overall landscape integrity in
Carlsbad Field Office planning area is at a moderate level. The resulting map identifies areas of
low and high landscape integrity in CFO, in which low landscape integrity may be attributed by
urban and industrial development, and agriculture. Results from the diversity metrics suggest that
landscape diversity remained relatively low from the time period of 2001 to 2016, peaking at
2013 and declining in later years. The linear regression results show a moderate positive and
significant correlation between the Landscape Integrity Index and Landscape Condition Map.
And the Welch’s two sample t-test results show that the mean LII value in protected areas is
slightly higher than the mean LII value in multiple-use areas.
4.1. The Landscape Integrity Index Model
The Landscape Integrity Index showed that the overall average of landscape integrity
value was 0.48 (SD = 0.05), indicating a moderate level of landscape integrity at the Carlsbad
Field Office planning area. The region was mostly characterized by LII values of 0.45 to 0.55
(~55% of the region had LII values of 0.45 to 0.5 and ~35% of the region had LII values of 0.5 to
0.55); less than 5% of the region had high LII values (>0.7) or low LII values (0.2 to 0.3) (Figure
6).

49

Figure 6. Histogram of LII values at Carlsbad Field Office
The resulting map showed areas of low landscape integrity near the major cities and the
northeast corner of CFO planning area; and areas of high landscape integrity were near central
and southwest corner of CFO planning area (Figure 7). There were very few areas with the
highest landscape integrity (LII values of 0.8 to 1), which were scattered throughout the region
and located at the CFO boundary, which could have resulted from the issue of using a boundary
constraint. Areas of moderately high landscape integrity (LII values of 0.5 to 0.7) occurred in
Arid Llano Estacado, Chihuahuan Desert Grasslands, Southern New Mexico Dissected Plains,
and Madrean Lower Montane Woodlands ecoregions. And areas of lowest landscape integrity
(LII values of 0 to 0.3) were located at areas of high human influence such as urban areas,
development areas, and agricultural areas (cultivated crops).
50

Figure 7. Map of Landscape Integrity Index at Carlsbad Field Office Planning Area (2001-2018).
LII values ranged from 0 (low landscape integrity) to 1 (high landscape integrity).
Maps of ecological integrity and landscape metrics showed that areas with low ecological
integrity and high landscape fragmentation occurred in high human influence areas (Figure 8).
Maps of resource- metrics and stressor-based metrics depicted that management actions for
vegetative communities were concentrated on central and southern CFO planning area whereas
management actions for oil and gas development were in central-north and central-east (Figure
8).
51

Figure 8. Maps of Ecological Integrity, Landscape Metrics, Resource-based Metrics, and
Stressor-based Metrics at Carlsbad Field Office Planning Area (2001-2018)
4.2. Landscape Diversity Metrics
The Diversity Metrics calculated using R suggested that landscape diversity was the
highest at 2013 and then declined in later years. Shannon’s Diversity Index and Simpson’s
Diversity Index both showed a similar trend of increased diversity from 2004 to 2013 and
reduced diversity from 2013 to 2016 (Figures 9a and 9b). The Simpson’s Diversity Index values
of 0.3 to 0.4 indicated that landscape diversity remained relatively low from the time period of
2001 to 2016. Values from the Shannon’s Evenness Index and Simpson’s Evenness Index were
both low, meaning that there was an uneven distribution of area among patch types (Figures 9c
52

and 9d). There were 14 patch types (i.e. the number of NLCD land cover classes) and 0.00055 as
the patch richness density, and these values remained constant throughout the time period.

a
0.733
0.73
0.756
0.765
0.828
0.821
0.807
0.72
0.74
0.76
0.78
0.8
0.82
0.84
2000 2002 2004 2006 2008 2010 2012 2014 2016 2018
Shannon’s Diversity Index (SHDI)
b
0.357 0.355
0.369
0.376
0.427
0.418
0.4
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
2000 2002 2004 2006 2008 2010 2012 2014 2016 2018
Simpson’s Diversity Index (SIDI)
53

Figure 9. Scatter Plots from 2001, 2004, 2006, 2008, 2011, 2013, and 2016 of
(a) Shannon’s Diversity Index (SHDI); (b) Simpson’s Diversity Index (SIDI);
(c) Shannon’s Evenness Index (SHEI); (d) Simpson’s Evenness Index (SIEI)
4.3. Model Validation Results
The first model validation process found a moderate positive and significant correlation
between 100 randomly selected LII values and the Landscape Condition Map (r = 0.5, p-value =
0.0000000189) (Table 7). The scatter plot and linear regression line of LII and LCM values
visualized the linear and positive relationship between LCM and LII (Figures 10a and 10b). The
box plot and density plot showed that both LCM and LII values contained several outliers, and
0.278
0.277
0.286
0.29
0.314
0.311
0.306
0.275
0.28
0.285
0.29
0.295
0.3
0.305
0.31
0.315
0.32
2000 2002 2004 2006 2008 2010 2012 2014 2016 2018
Shannon’s Evenness Index (SHEI)
c
0.384 0.382
0.398
0.405
0.46
0.45
0.341
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
2000 2002 2004 2006 2008 2010 2012 2014 2016 2018
Simpson’s Evenness Index (SIEI)
d
54

the distribution of LCM skewed left (towards high values) and LII values skewed right (towards
low values) (Figures 10c and 10d).
Table 7. Model Validation Results for Linear Regression Model
Model r r² Adjusted r² Std. Error of the Estimate p-value
LII and LCM 0.5261 0.2768 0.2694 0.0002 1.89e-08 *
* indicates significance
The linear regression results showed that the LCM could explain 28% of the correlation with LII
(r² = 0.28) (Table 7), which could be attributed to the inclusion of human land use change
datasets that reflected urban and industrial development, and managed and modified land cover.

LCM Value
LII Value
a
Scatter Plot of LII and LCM Values
Linear Regression Line of LII and LCM Values
LII Value
LCM Value
b
55

Figure 10. (a) Scatter Plot of LII and LCM Values;
(b) Linear Regression Line of LII and LCM Values;
(c) Box Plot of LII and LCM Values; (d) Density Plot of LII and LCM Values
Visually comparing the LII and LCM, the low landscape integrity areas both concentrated on the
urban areas, development areas, and agricultural areas (Figure 11).
c
d
56

Figure 11. Landscape Integrity Index Map and Landscape Condition Map
The second model validation process found that LII values in protected areas were
slightly higher than LII values in multiple-use areas (Figure 12) At a 95% confidence level, there
was a statistically significant difference between LII mean value in protected areas and LII mean
value in multiple-use areas (p-value = 0.03) (Table 8).

Figure 12. Box Plot of LII Values within Protected Areas and Multiple-Use Areas

0.7
0.6
0.5
Multiple-Use Protected
57

On average, the LII value within protected areas was 0.5, whereas the LII value within multiple-
use area was 0.49 (Table 8). Since there was only a small portion of protected areas in CFO
planning area and some of protected areas were also part of multiple-use areas, this might
explain why the mean values were very similar.
Table 8. Model Validation Results for Welch’s Two Sample T-Test
Model mean p-value
LII values in protected areas 0.4983 0.03295*
LII values in multiple-use areas 0.4887 0.03295*
* indicates significance under 95% confidence interval

58

Chapter 5 Discussion and Conclusions
This chapter discusses into the landscape integrity of CFO planning area and how LII can be
applied in the BLM cumulative effects analysis, including the Assessment, Inventory, and
Monitoring Strategy; decision-making process; and public communication and outreach. In this
final chapter, research limitations as data, analysis, and processing, as well as future research
opportunities to improve the methodology, LII approach, and communication and sharing of LII
model are discussed. Overall, the Landscape Integrity Index offers a comprehensive,
standardized, and transparent way to evaluate the cumulative impacts of BLM resource
management programs and can be used to improve the cumulative effects analysis.
5.1. Landscape Integrity of CFO Planning Area
The results of the LII model suggest that the overall landscape integrity of CFO planning
area is moderate, with low landscape integrity in urban and agricultural areas and high landscape
integrity near the central and southwest corner of CFO planning area. Low landscape integrity
can indicate low ecological integrity, low resources, high resource uses, high stressors, or high
landscape fragmentation, or all of the above on the ecosystem. While most of the region harbors
moderate levels of landscape integrity, this result represents a simplified view because only two
resource management programs were examined in this LII model. Nonetheless, the LII map
reveals valuable information for landscape-level planning even if this model may not have
captured all of the spatial complexities and relationships that exist in the region. However, these
findings provide direction for appropriate management of the landscape. Areas of substantial
human footprints and disturbance can become potential areas of restoration management for the
BLM, and/or should be further developed until landscape integrity is improved. Areas of high
59

landscape integrity are important zones for resources and ecosystem services that need active
monitoring to maintain their integrity.
To effectively manage multiple-use lands, certain LII values can be used as a limiting
factors for accepting or restricting proposed actions and as indicators for restoration planning.
For instance, land managers need to be vigilant in considering proposed actions that involve
stressor(s) on the ecosystem in areas with LII value of 0.3 or lower. These areas may require
restoration planning and/or extensive studies on resource presence and resources. Considering
that the mean LII value in protected areas is very similar to the mean LII value in multiple-use
areas, it may be necessary for land managers to inspect the type of activities that are allowed in
protected areas and identify areas of high conservation priority.
5.2. Applications of LII
The Bureau of Land Management can apply the Landscape Integrity Index in cumulative
effects analysis; Assessment, Inventory, and Monitoring (AIM) Strategy, decision-making
process; and communication and outreach to the public. For CEA, the land manager can develop
several LII models including baseline condition, with proposed actions, and with alternative
actions, and compare the LII values to evaluate the cumulative effects of the proposed actions
and alternatives. The land manager will have to identify thresholds of significance to determine
if there is a substantial cumulative impact. More steps are needed to conduct a thorough CEA,
but the addition of the LII can improve the CEA approach and produce pertinent information and
measures regarding the cumulative impacts of BLM resource management programs. Moreover,
the improvement of CEA process through the inclusion of scientifically sound data and credible
justification for data selection can lead to reduction in court challenges and court case losses.
60

The measures and metrics produced in the LII model will enhance the AIM Strategy by
determining ecosystem conditions and identifying potential monitoring locations. The LII model
can help establish the baseline condition using vegetation and land cover datasets and provide
precise values to express change and/or fluctuation in the condition of natural resources on
public lands. The LII model can also be used to investigate and prioritize specific areas of
concerns or areas of low landscape integrity, explore ecological restoration practices, measure
the effectiveness of those practices, and suggest potential monitoring locations in those areas.
The LII data layers and result values provide a summary of the cumulative impacts of the
BLM resource management programs, proposed actions, and alternatives to help inform the
decision-making process. The LII values and maps can help land managers and decision-makers
in examining areas of low and high landscape integrity and understand cumulative effects of the
programs and/or proposed actions. The comprehensive, transparent, and standardized GIS
approach provides appropriate data and credible justification for selection of data that will help
land managers make a better, more informed decision as to how to manag the public lands.
Finally, the LII map is a clear medium that can communicate to the public as to the
combined effects of BLM resource management programs actions on the environment. It can
improve public understanding of the multiple-use mission of the BLM and the landscape
condition of public lands.
5.3. Research Limitations
Although the LII model includes various ecological indicators, resource- and stressor-
based metrics, and landscape metrics, there are several data and analysis limitations imposed by
the lack of other field data and the limited R-functions for calculating landscape metrics. There
was no terrestrial data from the AIM data in Carlsbad Field Office Planning Area, which meant
61

that terrestrial core field measurements like bare ground, vegetation composition, vegetation
height, plant canopy gaps, non-native invasive plant species, and plant species of management
concern were not included as ecological indicators in the development of this LII model. Field
datasets that provide similar measurements from CFO were unable to obtained at this time, but
both AIM and field datasets can be used in future development of LII models. Some landscape
metrics such as proximity/isolation and contrast were not included in the development of this LII
model because the landscapemetrics package in R did not have the functions for calculating
those metrics. In the future, the authors of the package may add those functions or more research
can be done to use other landscape metrics as alternatives in identifying proximity and contrast.
Another data limitation is the data quality of the field data, in which inaccuracies in data
due to measurement error, data collection error, or human error will affect the precision and
completeness of the LII model. Moreover, missing data, duplicated data, and inaccurate data will
modify the cumulative effects at a spatial and temporal level by including or excluding
management actions. These problems need to be addressed by the data stewards and recognized
by the land managers or CEA practitioners who may use the field data in their analysis.
There was insufficient memory necessary to process complex statements for multiplying
multiple rasters and preserving all pixel values, which proved another limitation in the LII model.
Instead of only using the lowest LII value, a more comprehensive way to combine overlapping
stressors is to use either the summation approach (i.e. adding the values of the effects) or the
product-based approach (i.e. multiplying the values of the effects). The summation approach can
be used on additive interactions, and the product-based approach can be used on countervailing
and synergistic interactions. Future endeavors of enhancing the LII model should include the
62

summation and product-based approach, and identify less complex statements for adding or
multiplying multiple rasters or use a more powerful computer to process the complex statements.
5.4. Future Research Opportunities
The methodology presented in this study could benefit from expert opinions and research
on spatial and temporal scales, indicator species, site impact scores, and buffer distance for
resource- and stressor-based metrics. As mentioned earlier, the determination of spatial and
temporal scope should be based on the specific resource or ecosystem being impacted and the
duration of the effects. For example, a future development of the LII model to determine the
cumulative effects of a proposed oil and gas well should consider the resource(s) being impacted
by the proposed action and the duration of the effect (e.g. clearing the land, drilling the well,
extracting oil or gas from the well, and burying the well) to define the spatial and temporal
scopes of this project. With additional expert opinions from BLM staff, the selection of indicator
species can be expanded to include more pertinent BLM management species and identify acres
and distance measurements for assigning weights to the patch size and structural connectivity
variables. With the inclusion of other species, the connectivity variable requires additional
review and potentially new representation because it is fluid and varies amongst species.
Adjusting the site impact scores of ecological integrity indicators and resource- and stressor-
based metrics though expert opinions can be another future progression of the LII model, adding
more credible justification for selection of data and exploring the range of LII values. A future
addition to enhance the LII model is identifying the impact radius and creating buffers for
resource- and stressor-based metrics. For instance, the existing oil and gas wells can be buffered
using a distance of 49.47 m to create a zone of 7689.03 m² to simulate the approximate area of
surface disturbance caused by the wells.
63

Future directions to strengthen the LII approach include integrating additional steps into
the three-stage workflow, refining the indicators and metrics, and investigating other ways to
weight the landscape metrics and examine landscape diversity metrics. In the first stage of the
workflow, mechanisms such as target resources, key stressors, societally desired conditions, and
thresholds of significance can be included in the assessment report to inform management. In the
second stage of the workflow, BLM staff can define the natural reference and societally desired
conditions and analyze the deviation of the LII value resulted from the management action or
proposed action. In the third stage of the workflow, establishing thresholds of significance will
help with informing management by delineating what management actions to take if the LII
value reaches a certain number or what it means to have a low or high landscape integrity. Future
assessments should consider a wide range of indicators and metrics that would encompass
priority resources, ecosystem services, and sub-surface disturbance, depending on the mission
and region of the BLM office conducting the LII analysis. In this analysis, the landscape metrics
were normalized from 0 to 1 given the range or the minimum and maximum of the results. To
better represent the significance of the landscape metrics, there should be a deeper look at other
classification or weighting options for ranking the landscape metrics from 0 to 1. And even
though landscape diversity metrics could not be included in the development of the LII model, its
decreasing trend revealed landscape patterns that could be worthwhile to examine. Future
research direction can tackle the complex question of which management decisions could have
caused the reduction in landscape diversity.
Future efforts in communication and sharing of the LII model could create ArcGIS
StoryMaps as a public relations outreach medium, convert the python code from Python 2.7 to
Python 3 for ArcGIS Pro use, and collaborate with other BLM offices to compare the selection
64

of data and weighting options for indicators and metrics. In addition to Resource Management
Plans, ArcGIS StoryMaps can illustrate the cumulative effects of BLM resource management
programs or proposed actions via interactive visuals of the indicators, metrics, and the LII map.
Incorporating LII StoryMaps in the BLM website may help the public gain a better
understanding of what is happening on public lands and create awareness for the multiple-use
and sustained yield mission of BLM. With the migration of ArcMap to ArcGIS Pro, the python
script will need to be upgraded to Python 3 to access the ArcGIS Pro functionalities and
geoprocessing tools. The advancement of LII model also relies on the coordination between
BLM offices, where neighboring offices can share field data, expert opinions on selection of data
and composite scoring system, and LII results for comparison.
5.5. Conclusion
The Bureau of Land Management is constantly striving to balance the complex multiple-
use and sustained yield mission of protecting the resources of our public lands and generating
revenue through development. The short- and long-term impacts of the diverse range of BLM
resource management programs and proposed actions are affecting the landscape in various
ways, and it is essential to understand and analyze these cumulative impacts so that we can
manage the public lands in a sustainable manner. The Landscape Integrity Index evaluates the
cumulative effects of these programs by using indicators and metrics to examine the ecological
integrity, resources, resource uses, stressors, and landscape patterns and relationships. The GIS
approach proposed in this study builds on the cumulative effects analysis and evaluation of
ecological integrity to assess the landscape condition in a comprehensive, standardized, and
transparent process. The GIS model of LII with python scripts and R Markdown will be available
to download at https://github.com/liling2lee/Landscape_Integrity_Index. With future
65

improvements made to the LII model, it will address management goals and objectives,
incorporate relevant and pertinent indicators and metrics, and facilitate planning and
management across BLM offices.
66

References
Allred, Brady W., W. Kolby Smith, Dirac Twidwell, Julia H. Haggerty, Steven W. Running,
David E. Naugle, and Samuel D. Fuhlendorf. 2015. “Ecosystem Services Lost to Oil and
Gas in North America.” Science 348 (6233): 401–2.
https://doi.org/10.1126/science.aaa4785.
AmphibiaWeb. 2019. University of California, Berkeley, CA, USA. Accessed October 13, 2019.
http://amphibiaweb.org.
Andreasen, James K, Robert V O’Neill, Reed Noss, and Nicholas C Slosser. 2001.
“Considerations for the Development of a Terrestrial Index of Ecological Integrity.”
Ecological Indicators 1 (1): 21–35. https://doi.org/10.1016/S1470-160X(01)00007-3.
Animal Diversity Web. 2014. University of Michigan, Museum of Zoology, Ann Arbor, MI,
USA. Accessed October 13, 2019. https://animaldiversity.org/.
Atkinson, Samuel F., and Larry W. Canter. 2011. “Assessing the Cumulative Effects of Projects
Using Geographic Information Systems.” Environmental Impact Assessment Review,
SPECIAL ISSUE on Cumulative Effects Assessment, 31 (5): 457–64.
https://doi.org/10.1016/j.eiar.2011.01.008.
Bureau of Land Management, Office of the Assistant Director, Renewable Resources and
Planning. “BLM NEPA Handbook H-1790-1.” U.S. Washington, D.C.: Bureau of Land
Management Office of the Assistant Director, Renewable Resources and Planning, 2008.
https://www.ntc.blm.gov/krc/uploads/366/NEPAHandbook_H-1790_508.pdf.
Bureau of Land Management. 2018a. “NM Sensitive Update: Amphibians, arthropods, birds,
crustaceans, fish, mammals, mollucks, and reptiles.” Accessed October 13, 2019.
https://www.blm.gov/sites/blm.gov/files/IB%20NM-2019-
002_Attachment%201%20Animal.pdf
Bureau of Land Management, New Mexico State Office. “Draft Resource Management Plan and
Environmental Impact Statement.” New Mexico: Bureau of Land Management, New
Mexico State Office, 2018b. https://eplanning.blm.gov/epl-front-
office/projects/lup/64444/153042/187358/BLM_CFO_Draft_RMP_-_Volume_I_-_EIS_-
_August_2018_(1).pdf.
Canter, Larry W., and J. Kamath. 1995. “Questionnaire Checklist for Cumulative Impacts.”
Environmental Impact Assessment Review 15 (4): 311–39. https://doi.org/10.1016/0195-
9255(95)00010-C.
Canter, Larry W., and S. F. Atkinson. 2011. “Multiple-uses of Indicators and Indices in
Cumulative Effects Assessment and Management.” Environmental Impact Assessment
Review 31 (5): 491–501. https://doi.org/10.1016/j.eiar.2011.01.012.
67

Carter, Sarah K., Erica Fleishman, Ian I. F. Leinwand, Curtis H. Flather, Natasha B. Carr, Frank
A. Fogarty, Matthias Leu, Barry R. Noon, Martha E. Wohlfeil, and David J. A. Wood.
2019. “Quantifying Ecological Integrity of Terrestrial Systems to Inform Management of
Multiple-Use Public Lands in the United States.” Environmental Management, April.
https://doi.org/10.1007/s00267-019-01163-w.
Copeland, Holly E., Kevin E. Doherty, David E. Naugle, Amy Pocewicz, and Joseph M.
Kiesecker. 2009. “Mapping Oil and Gas Development Potential in the US Intermountain
West and Estimating Impacts to Species.” PLOS ONE 4 (10): e7400.
https://doi.org/10.1371/journal.pone.0007400.
Crain, Caitlin Mullan, Kristy Kroeker, and Benjamin S. Halpern. 2008. “Interactive and
Cumulative Effects of Multiple Human Stressors in Marine Systems.” Ecology Letters 11
(12): 1304–15. https://doi.org/10.1111/j.1461-0248.2008.01253.x.
Dale, Virginia H., and Suzanne C. Beyeler. 2001. “Challenges in the Development and Use of
Ecological Indicators.” Ecological Indicators 1 (1): 3–10. https://doi.org/10.1016/S1470-
160X(01)00003-6.
Esri. 2019a. “Euclidean distance.” Environmental Systems Research Institute. Accessed August
25, 2019. http://desktop.arcgis.com/ja/arcmap/10.6/tools/spatial-analyst-
toolbox/euclidean-distance.htm.
____. 2019b. “Region Group.” Environmental Systems Research Institute. Accessed August 25,
2019. http://desktop.arcgis.com/ja/arcmap/10.6/tools/spatial-analyst-toolbox/region-
group.htm.
Foley, Melissa M., Lindley A. Mease, Rebecca G. Martone, Erin E. Prahler, Tiffany H.
Morrison, Cathryn Clarke Murray, and Deborah Wojcik. 2017. “The Challenges and
Opportunities in Cumulative Effects Assessment.” Environmental Impact Assessment
Review 62: 122–34. https://doi.org/10.1016/j.eiar.2016.06.008.
Frazier, A. 2019. Landscape Metrics. The Geographic Information Science & Technology Body
of Knowledge. Edited by John P. Wilson. 2nd ed. DOI: 10.22224/gistbok/2019.2.3.
Hak, John C., and Patrick J. Comer. 2017. “Modeling Landscape Condition for Biodiversity
Assessment—Application in Temperate North America.” Ecological Indicators 82: 206–
16. https://doi.org/10.1016/j.ecolind.2017.06.049.
Harvey, Bret C., and Steven F. Railsback. 2007. “Estimating Multi-Factor Cumulative
Watershed Effects on Fish Populations with an Individual-Based Model.” Fisheries 32
(6): 292–98. https://doi.org/10.1577/1548-8446(2007)32[292:EFCWEO]2.0.CO;2.
Hobbs, Richard J., David N. Cole, Laurie Yung, Erika S. Zavaleta, Gregory H. Aplet, F. Stuart
Chapin, Peter B. Landres, et al. 2010. “Guiding Concepts for Park and Wilderness
Stewardship in an Era of Global Environmental Change.” Frontiers in Ecology and the
Environment. 8(9): 483-490 8 (9): 483–90.
68

Homer, C.G., J.A. Dewitz, L. Yang, S. Jin, P. Danielson, G. Xian, J. Coulston, N.D. Herold, J.D.
Wickham, and K. Megown. 2015. “Completion of the 2011 National Land Cover
Database for the conterminous United States-Representing a decade of land cover change
information.” Photogrammetric Engineering and Remote Sensing 81 (5): 345-354.
https://cran.r-project.org/web/packages/landscapemetrics/landscapemetrics.pdf.
IUCN. 2019. The IUCN Red List of Threatened Species. Accessed October 13, 2019.
https://www.iucnredlist.org.
Landscape Fire and Resource Management Planning Tools (LANDFIRE). 2016a. “LANDFIRE
Program: Data Products - Fire Regime - Vegetation Departure.” U.S. Geologic Survey.
Accessed July 21, 2019. https://www.landfire.gov/vdep.php.
______. 2016b. “LANDFIRE Program: Data Products - Vegetation - Existing Vegetation Type.”
U.S. Geologic Survey. Accessed July 21, 2019. https://www.landfire.gov/evt.php.
McGarigal, Kevin, and Barbara J. Marks. 1995. “FRAGSTATS: Spatial Pattern Analysis
Program for Quantifying Landscape Structure.” Gen. Tech. Rep. PNW-GTR-351.
Portland, OR: U.S. Department of Agriculture, Forest Service, Pacific Northwest
Research Station. https://doi.org/10.2737/PNW-GTR-351.
McGarigal, Kevin. 2015. “FRAGSTATS HELP.” LandEco Consulting.
https://www.umass.edu/landeco/research/fragstats/documents/fragstats.help.4.2.pdf
McGarigal, Kevin, Bradley W. Compton, Ethan B. Plunkett, William V. DeLuca, Joanna Grand,
Eduard Ene, and Scott D. Jackson. 2018. “A Landscape Index of Ecological Integrity to
Inform Landscape Conservation.” Landscape Ecology 33 (7): 1029–48.
https://doi.org/10.1007/s10980-018-0653-9.
Natural Heritage New Mexico. n.d. “New Mexico Rare Plant Conservation Strategy | NHNM.”
Natural Heritage New Mexico. Accessed July 21, 2019.
https://nhnm.unm.edu/botany/nm_rare_plant_conservation_strategy.
New Mexico Energy, Minerals, and Natural Resources Department (EMNRD) Forestry Division.
New Mexico Rare Plant Conservation Strategy Partnership. New Mexico Rare Plant
Conservation Strategy. Open-file report. Santa Fe, New Mexico, 2017.
http://www.emnrd.state.nm.us/SFD/documents/NM%20Rare%20Plant%20Conservation
%20Strategy_03202019.pdf.
Parrish, Jeffrey D., David P. Braun, and Robert S. Unnasch. 2003. “Are We Conserving What
We Say We Are? Measuring Ecological Integrity within Protected Areas.” BioScience 53
(9): 851–60. https://doi.org/10.1641/0006-3568(2003)053[0851:AWCWWS]2.0.CO;2.
Richardson, Leif L., Kent P. McFarland, Sara Zahendra, and Spencer Hardy. 2019. “Bumble bee
(Bombus) distribution and diversity in Vermont, USA: a century of change. Journal of
Insect Conservation, 23 (1): 45-62. https://doi.org/10.1007/s10841-018-0113-5.
69

Smith, Michael D. 2006. “Research Article: Cumulative Impact Assessment under the National
Environmental Policy Act: An Analysis of Recent Case Law.” Environmental Practice 8
(4): 228–40. https://doi.org/10.1017/S1466046606060467.
Smithsonian. 2019. “Animals.” Smithsonian’s National Zoo & Conservation Biology Institute,
Washington, DC, USA. Accessed October 13, 2019. https://nationalzoo.si.edu/animals.
The Cornell Lab of Ornithology. 2019a. “All About Birds.” Cornell University, Ithaca, NY,
USA. Accessed October 13, 2019. https://www.allaboutbirds.org/guide/.
The Cornell Lab of Ornithology. 2019b. “Birds of North America.” Cornell University, Ithaca,
NY, USA. Accessed October 13, 2019. https://birdsna.org/Species-Account/bna/home.
Spencer, David, David Haukos, Christian Hagen, Melinda Daniels, and Doug Goodin. 2017.
“Conservation Reserve Program Mitigates Grassland Loss in the Lesser Prairie-Chicken
Range of Kansas.” Global Ecology and Conservation 9: 21–38.
https://doi.org/10.1016/j.gecco.2016.11.004.
Theobald, David M. 2013. “A General Model to Quantify Ecological Integrity for Landscape
Assessments and US Application.” Landscape Ecology; Dordrecht 28 (10): 1859–74.
http://dx.doi.org.libproxy2.usc.edu/10.1007/s10980-013-9941-6.
U.S. Council of Environmental Quality. Considering Cumulative Effects Under the National
Environmental Policy Act. U.S. Washington, D.C.: U.S. Council of Environmental
Quality, 1997.
https://www.energy.gov/sites/prod/files/nepapub/nepa_documents/RedDont/G-CEQ-
ConsidCumulEffects.pdf.
U.S. Department of the Interior. Bureau of Land Management, Office of the Solicitor (editors).
The Federal Land Policy and Management Act, as amended. U.S. Washington, D.C.:
Department of the Interior, Bureau of Land Management Office of Public Affairs, 2001.
https://www.blm.gov/or/regulations/files/FLPMA.pdf.
U.S. Department of the Interior. U.S. Geologic Survey. Multiscale Guidance and Tools for
Implementing a Landscape Approach to Resource Management in the Bureau of Land
Management, by Carter, Sarah K., Natasha B. Carr, Kevin H. Miller, and David J.A.
Wood. Open-file report 2016–1207. Reston, Virginia, 2017.
http://pubs.er.usgs.gov/publication/ofr20161207.
U.S. Fish & Wildlife Service. 2019. “Featured Pollinator.” Accessed October 13, 2019.
https://www.fws.gov/pollinators/features/Monarch_Butterfly.html.
Walston, Leroy J., and Heidi M. Hartmann. 2018. “Development of a Landscape Integrity Model
Framework to Support Regional Conservation Planning.” PLOS ONE 13 (4): e0195115.
https://doi.org/10.1371/journal.pone.0195115.
70

Wang, Xianli, F. Guillaume Blanchet, and Nicola Koper. 2014. “Measuring Habitat
Fragmentation: An Evaluation of Landscape Pattern Metrics.” Methods in Ecology and
Evolution 5 (7): 634–46. https://doi-org.libproxy1.usc.edu/10.1111/2041-210X.12198.
Western Association of Fish and Wildlife Agencies. The Lesser Prairie-Chicken Range-wide
Conservation Plan, by Van Pelt, William E., S. Kyle, J. Pitman, D. Klute, G. Beauprez,
D. Schoeling, A. Janus, and J. Haufler. Open-file report. Cheyenne, Wyoming, 2013.
https://www.wafwa.org/Documents%20and%20Settings/37/Site%20Documents/Initiative
s/Lesser%20Prairie%20Chicken/2013LPCRWPfinalfor4drule12092013.pdf.
Wu, Zhenhua, Shaogang Lei, Bao-Jie He, Zhengfu Bian, Yinghong Wang, Qingqing Lu,
Shangui Peng, and Linghua Duo. 2019. “Assessment of Landscape Ecological Health: A
Case Study of a Mining City in a Semi-Arid Steppe.” International Journal of
Environmental Research and Public Health 16 (5).
https://doi.org/10.3390/ijerph16050752.
Wurtzebach, Zachary, and Courtney Schultz. 2016. “Measuring Ecological Integrity: History,
Practical Applications, and Research Opportunities.” BioScience 66 (6): 446–57.
https://doi.org/10.1093/biosci/biw037.
71

Appendix A Management Indicator Species for Carlsbad Field Office, NM
Species Common
Name
Taxonomic
Group
NM Status BLM
Status
Habitat
(Vegetation
Area)
Craugastor
(Eleutherodactylus)
augustilatrans

Eastern
barking frog
Amphibian SGCN Watch Scrub
(Shrubland)
Gastrophryne
olivacea
Western
narrowmouth
toad
Amphibian Endangered,
SGCN
Watch Grasslands
Lithobates (Rana)
blairi
Plains
leopard frog
Amphibian SGCN Watch Grasslands
Danaus plexippus
plexippus
Monarch
Butterfly
Arthropods SGCN Watch
*New*
Shrubland
Bombus
occidentalis
Western
Bumble Bee
Arthropods None Watch
*New*
Shrubland
Athene cunicularia Western
Burrowing
Owl
Birds SGCN BLM
Sensitive
Grasslands
Anthus spragueii Sprague's
Pipit
Birds SGCN BLM
Sensitive
Grasslands
Calcarius
mccownii
McCown's
Longspur
Birds SGCN BLM
Sensitive
*New*
Grasslands
Calcarius ornatus Chestnut-
collared
Longspur
Birds SGCN BLM
Sensitive
Grasslands
Tympanuchus
pallidicinctus
Lesser
Prairie-
chicken
Birds SGCN BLM
Sensitive
Grasslands
Vireo belLII
arizonae
Bell's Vireo Birds Threatened,
SGCN
BLM
Sensitive
Scrub
(Shrubland)
Vermivora
virginiae
Virginia's
Warbler
Birds SGCN BLM
Sensitive
*New*
Open
Woodlands
(Conifer-
Hardwood)
Aphelocoma
woodhouseii
Woodhouse's
Scrub- Jay
Birds None Watch
*New*
Scrub
(Shrubland)
Aquila chrysaetos Golden Eagle Birds None Watch Grasslands
Botaurus
lentiginosus
American
Bittern
Birds SGCN Watch Marshes
(Riparian)
Buteogallus
anthracinus
Common
Black-Hawk
Birds Threatened,
SGCN
Watch Riparian
72

Callipepla
squamata
Scaled Quail Birds None Watch
*New*
Grasslands
Carpodacus
cassinii
Cassin's
Finch
Birds SGCN Watch Forests
(Conifer)
Lanius
ludovicianus
Loggerhead
Shrike
Birds SGCN Watch Open
Woodlands
(Shrubland)
Melanerpes lewis Lewis's
Woodpecker
Birds SGCN Watch
*New*
Open
Woodlands
(Conifer)
Numenius
americanus
Long- billed
Curlew
Birds SGCN Watch Grasslands
Oreoscoptes
montanus
Sage
Thrasher
Birds None Watch Scrub
(Shrubland)
Passerina ciris Painted
Bunting
Birds None Watch Scrub
(Shrubland)
Spizella atrogularis
evura
Black-
chinned
Sparrow
Birds SGCN Watch Scrub
(Shrubland)
Vireo vicinior Gray Vireo Birds Threatened,
SGCN
Watch Scrub
(Shrubland)
Corynorhinus
townsendii
Townsend's
big-eared bat
Mammals SGCN BLM
Sensitive
Forest (Conifer)
Cynomys
ludovicianus
Black-tailed
prairie dog
Mammals SGCN BLM
Sensitive
Grasslands
Cratogeomys
castanops
Yellow-faced
pocket
gopher
Mammals None Watch Grasslands
Cryptotis parva Least shrew Mammals Threatened,
SGCN
Watch
*New*
Grasslands
Nyctinomops
femorosaccus
Pocketed
free-tailed bat
Mammals None Watch Desertlands
(Shrubland)
Sistrurus
tergeminus
Desert
massasauga
Reptiles SGCN BLM
Sensitive
*New*
Desert
grasslands
(Grasslands)
Crotalus lepidus
lepidus
Mottled Rock
Rattlesnake
Reptiles Threatened,
SGCN
Watch Grasslands
Source: AmphibiaWeb (2019), Animal Diversity Web (2014), Bureau of Land Management
(2018a), IUCN 2019, Richardson et al. (2019), Smithsonian (2019), The Cornell Lab of
Ornithology (2019a), The Cornell Lab of Ornithology (2019b), and USFWS (2019)

73

Appendix B Landscape Metrics Results
Shape Metrics: Perimeter-Area Fractal Dimension (PAFRAC) – 2001
NLCD Class NLCD Class Value Range
11 Open Water 1.35 1 ≤ PAFRAC ≤ 2
21 Developed, Open Space 1.60 1 ≤ PAFRAC ≤ 2
22 Developed, Low Intensity 1.58 1 ≤ PAFRAC ≤ 2
23 Developed, Medium Intensity 1.60 1 ≤ PAFRAC ≤ 2
24 Developed, High Intensity 1.50 1 ≤ PAFRAC ≤ 2
31 Barren Land 1.39 1 ≤ PAFRAC ≤ 2
41 Deciduous Forest 1.76 1 ≤ PAFRAC ≤ 2
42 Evergreen Forest 1.52 1 ≤ PAFRAC ≤ 2
52 Shrub/Scrub 1.55 1 ≤ PAFRAC ≤ 2
71 Herbaceous 1.63 1 ≤ PAFRAC ≤ 2
81 Hay/Pasture 1.38 1 ≤ PAFRAC ≤ 2
82 Cultivated Crops 1.27 1 ≤ PAFRAC ≤ 2
90 Woody Wetlands 1.51 1 ≤ PAFRAC ≤ 2
95 Emergent Herbaceous Wetlands 1.50 1 ≤ PAFRAC ≤ 2

Shape Metrics: Perimeter-Area Fractal Dimension (PAFRAC) – 2004
NLCD Class NLCD Class Value Range
11 Open Water 1.37 1 ≤ PAFRAC ≤ 2
21 Developed, Open Space 1.60 1 ≤ PAFRAC ≤ 2
22 Developed, Low Intensity 1.58 1 ≤ PAFRAC ≤ 2
23 Developed, Medium Intensity 1.60 1 ≤ PAFRAC ≤ 2
24 Developed, High Intensity 1.50 1 ≤ PAFRAC ≤ 2
31 Barren Land 1.40 1 ≤ PAFRAC ≤ 2
41 Deciduous Forest 1.75 1 ≤ PAFRAC ≤ 2
42 Evergreen Forest 1.52 1 ≤ PAFRAC ≤ 2
52 Shrub/Scrub 1.55 1 ≤ PAFRAC ≤ 2
71 Herbaceous 1.63 1 ≤ PAFRAC ≤ 2
81 Hay/Pasture 1.40 1 ≤ PAFRAC ≤ 2
82 Cultivated Crops 1.26 1 ≤ PAFRAC ≤ 2
90 Woody Wetlands 1.51 1 ≤ PAFRAC ≤ 2
95 Emergent Herbaceous Wetlands 1.50 1 ≤ PAFRAC ≤ 2

74

Shape Metrics: Perimeter-Area Fractal Dimension (PAFRAC) – 2006
NLCD Class NLCD Class Value Range
11 Open Water 1.37 1 ≤ PAFRAC ≤ 2
21 Developed, Open Space 1.58 1 ≤ PAFRAC ≤ 2
22 Developed, Low Intensity 1.58 1 ≤ PAFRAC ≤ 2
23 Developed, Medium Intensity 1.60 1 ≤ PAFRAC ≤ 2
24 Developed, High Intensity 1.50 1 ≤ PAFRAC ≤ 2
31 Barren Land 1.40 1 ≤ PAFRAC ≤ 2
41 Deciduous Forest 1.72 1 ≤ PAFRAC ≤ 2
42 Evergreen Forest 1.52 1 ≤ PAFRAC ≤ 2
52 Shrub/Scrub 1.55 1 ≤ PAFRAC ≤ 2
71 Herbaceous 1.63 1 ≤ PAFRAC ≤ 2
81 Hay/Pasture 1.37 1 ≤ PAFRAC ≤ 2
82 Cultivated Crops 1.26 1 ≤ PAFRAC ≤ 2
90 Woody Wetlands 1.51 1 ≤ PAFRAC ≤ 2
95 Emergent Herbaceous Wetlands 1.51 1 ≤ PAFRAC ≤ 2

Shape Metrics: Perimeter-Area Fractal Dimension (PAFRAC) – 2008
NLCD Class NLCD Class Value Range
11 Open Water 1.38 1 ≤ PAFRAC ≤ 2
21 Developed, Open Space 1.58 1 ≤ PAFRAC ≤ 2
22 Developed, Low Intensity 1.58 1 ≤ PAFRAC ≤ 2
23 Developed, Medium Intensity 1.60 1 ≤ PAFRAC ≤ 2
24 Developed, High Intensity 1.50 1 ≤ PAFRAC ≤ 2
31 Barren Land 1.40 1 ≤ PAFRAC ≤ 2
41 Deciduous Forest 1.69 1 ≤ PAFRAC ≤ 2
42 Evergreen Forest 1.52 1 ≤ PAFRAC ≤ 2
52 Shrub/Scrub 1.54 1 ≤ PAFRAC ≤ 2
71 Herbaceous 1.63 1 ≤ PAFRAC ≤ 2
81 Hay/Pasture 1.37 1 ≤ PAFRAC ≤ 2
82 Cultivated Crops 1.26 1 ≤ PAFRAC ≤ 2
90 Woody Wetlands 1.51 1 ≤ PAFRAC ≤ 2
95 Emergent Herbaceous Wetlands 1.50 1 ≤ PAFRAC ≤ 2

Shape Metrics: Perimeter-Area Fractal Dimension (PAFRAC) – 2011
NLCD Class NLCD Class Value Range
11 Open Water 1.38 1 ≤ PAFRAC ≤ 2
21 Developed, Open Space 1.55 1 ≤ PAFRAC ≤ 2
75

22 Developed, Low Intensity 1.57 1 ≤ PAFRAC ≤ 2
23 Developed, Medium Intensity 1.60 1 ≤ PAFRAC ≤ 2
24 Developed, High Intensity 1.49 1 ≤ PAFRAC ≤ 2
31 Barren Land 1.41 1 ≤ PAFRAC ≤ 2
41 Deciduous Forest 1.66 1 ≤ PAFRAC ≤ 2
42 Evergreen Forest 1.52 1 ≤ PAFRAC ≤ 2
52 Shrub/Scrub 1.54 1 ≤ PAFRAC ≤ 2
71 Herbaceous 1.62 1 ≤ PAFRAC ≤ 2
81 Hay/Pasture 1.42 1 ≤ PAFRAC ≤ 2
82 Cultivated Crops 1.26 1 ≤ PAFRAC ≤ 2
90 Woody Wetlands 1.51 1 ≤ PAFRAC ≤ 2
95 Emergent Herbaceous Wetlands 1.50 1 ≤ PAFRAC ≤ 2

Shape Metrics: Perimeter-Area Fractal Dimension (PAFRAC) – 2013
NLCD Class NLCD Class Value Range
11 Open Water 1.40 1 ≤ PAFRAC ≤ 2
21 Developed, Open Space 1.55 1 ≤ PAFRAC ≤ 2
22 Developed, Low Intensity 1.57 1 ≤ PAFRAC ≤ 2
23 Developed, Medium Intensity 1.60 1 ≤ PAFRAC ≤ 2
24 Developed, High Intensity 1.49 1 ≤ PAFRAC ≤ 2
31 Barren Land 1.40 1 ≤ PAFRAC ≤ 2
41 Deciduous Forest 1.76 1 ≤ PAFRAC ≤ 2
42 Evergreen Forest 1.52 1 ≤ PAFRAC ≤ 2
52 Shrub/Scrub 1.54 1 ≤ PAFRAC ≤ 2
71 Herbaceous 1.62 1 ≤ PAFRAC ≤ 2
81 Hay/Pasture 1.39 1 ≤ PAFRAC ≤ 2
82 Cultivated Crops 1.26 1 ≤ PAFRAC ≤ 2
90 Woody Wetlands 1.51 1 ≤ PAFRAC ≤ 2
95 Emergent Herbaceous Wetlands 1.50 1 ≤ PAFRAC ≤ 2

Shape Metrics: Perimeter-Area Fractal Dimension (PAFRAC) – 2016
NLCD Class NLCD Class Value Range
11 Open Water 1.38 1 ≤ PAFRAC ≤ 2
21 Developed, Open Space 1.55 1 ≤ PAFRAC ≤ 2
22 Developed, Low Intensity 1.57 1 ≤ PAFRAC ≤ 2
23 Developed, Medium Intensity 1.59 1 ≤ PAFRAC ≤ 2
24 Developed, High Intensity 1.47 1 ≤ PAFRAC ≤ 2
31 Barren Land 1.43 1 ≤ PAFRAC ≤ 2
76

41 Deciduous Forest 1.67 1 ≤ PAFRAC ≤ 2
42 Evergreen Forest 1.52 1 ≤ PAFRAC ≤ 2
52 Shrub/Scrub 1.54 1 ≤ PAFRAC ≤ 2
71 Herbaceous 1.62 1 ≤ PAFRAC ≤ 2
81 Hay/Pasture 1.38 1 ≤ PAFRAC ≤ 2
82 Cultivated Crops 1.27 1 ≤ PAFRAC ≤ 2
90 Woody Wetlands 1.52 1 ≤ PAFRAC ≤ 2
95 Emergent Herbaceous Wetlands 1.51 1 ≤ PAFRAC ≤ 2

Core Area Metrics: Coefficient of Variation of Core Area Index (CAI_CV) – 2001
NLCD Class NLCD Class Value Range
11 Open Water 200 CAI_CV ≥ 0
21 Developed, Open Space 383 CAI_CV ≥ 0
22 Developed, Low Intensity 652 CAI_CV ≥ 0
23 Developed, Medium Intensity 560 CAI_CV ≥ 0
24 Developed, High Intensity 442 CAI_CV ≥ 0
31 Barren Land 215 CAI_CV ≥ 0
41 Deciduous Forest 332 CAI_CV ≥ 0
42 Evergreen Forest 162 CAI_CV ≥ 0
52 Shrub/Scrub 238 CAI_CV ≥ 0
71 Herbaceous 219 CAI_CV ≥ 0
81 Hay/Pasture 107 CAI_CV ≥ 0
82 Cultivated Crops 96 CAI_CV ≥ 0
90 Woody Wetlands 230 CAI_CV ≥ 0
95 Emergent Herbaceous Wetlands 246 CAI_CV ≥ 0

Core Area Metrics: Coefficient of Variation of Core Area Index (CAI_CV) – 2004
NLCD Class NLCD Class Value Range
11 Open Water 238 CAI_CV ≥ 0
21 Developed, Open Space 383 CAI_CV ≥ 0
22 Developed, Low Intensity 652 CAI_CV ≥ 0
23 Developed, Medium Intensity 560 CAI_CV ≥ 0
24 Developed, High Intensity 442 CAI_CV ≥ 0
31 Barren Land 216 CAI_CV ≥ 0
41 Deciduous Forest 336 CAI_CV ≥ 0
42 Evergreen Forest 162 CAI_CV ≥ 0
52 Shrub/Scrub 238 CAI_CV ≥ 0
71 Herbaceous 220 CAI_CV ≥ 0
77

81 Hay/Pasture 106 CAI_CV ≥ 0
82 Cultivated Crops 93 CAI_CV ≥ 0
90 Woody Wetlands 234 CAI_CV ≥ 0
95 Emergent Herbaceous Wetlands 270 CAI_CV ≥ 0

Core Area Metrics: Coefficient of Variation of Core Area Index (CAI_CV) – 2006
NLCD Class NLCD Class Value Range
11 Open Water 239 CAI_CV ≥ 0
21 Developed, Open Space 317 CAI_CV ≥ 0
22 Developed, Low Intensity 663 CAI_CV ≥ 0
23 Developed, Medium Intensity 582 CAI_CV ≥ 0
24 Developed, High Intensity 465 CAI_CV ≥ 0
31 Barren Land 213 CAI_CV ≥ 0
41 Deciduous Forest 336 CAI_CV ≥ 0
42 Evergreen Forest 162 CAI_CV ≥ 0
52 Shrub/Scrub 240 CAI_CV ≥ 0
71 Herbaceous 220 CAI_CV ≥ 0
81 Hay/Pasture 121 CAI_CV ≥ 0
82 Cultivated Crops 93 CAI_CV ≥ 0
90 Woody Wetlands 252 CAI_CV ≥ 0
95 Emergent Herbaceous Wetlands 277 CAI_CV ≥ 0

Core Area Metrics: Coefficient of Variation of Core Area Index (CAI_CV) – 2008
NLCD Class NLCD Class Value Range
11 Open Water 257 CAI_CV ≥ 0
21 Developed, Open Space 317 CAI_CV ≥ 0
22 Developed, Low Intensity 663 CAI_CV ≥ 0
23 Developed, Medium Intensity 582 CAI_CV ≥ 0
24 Developed, High Intensity 465 CAI_CV ≥ 0
31 Barren Land 212 CAI_CV ≥ 0
41 Deciduous Forest 340 CAI_CV ≥ 0
42 Evergreen Forest 162 CAI_CV ≥ 0
52 Shrub/Scrub 241 CAI_CV ≥ 0
71 Herbaceous 220 CAI_CV ≥ 0
81 Hay/Pasture 129 CAI_CV ≥ 0
82 Cultivated Crops 91 CAI_CV ≥ 0
90 Woody Wetlands 235 CAI_CV ≥ 0
95 Emergent Herbaceous Wetlands 285 CAI_CV ≥ 0
78

Core Area Metrics: Coefficient of Variation of Core Area Index (CAI_CV) – 2011
NLCD Class NLCD Class Value Range
11 Open Water 238 CAI_CV ≥ 0
21 Developed, Open Space 288 CAI_CV ≥ 0
22 Developed, Low Intensity 682 CAI_CV ≥ 0
23 Developed, Medium Intensity 561 CAI_CV ≥ 0
24 Developed, High Intensity 457 CAI_CV ≥ 0
31 Barren Land 215 CAI_CV ≥ 0
41 Deciduous Forest 355 CAI_CV ≥ 0
42 Evergreen Forest 165 CAI_CV ≥ 0
52 Shrub/Scrub 238 CAI_CV ≥ 0
71 Herbaceous 224 CAI_CV ≥ 0
81 Hay/Pasture 126 CAI_CV ≥ 0
82 Cultivated Crops 91 CAI_CV ≥ 0
90 Woody Wetlands 241 CAI_CV ≥ 0
95 Emergent Herbaceous Wetlands 274 CAI_CV ≥ 0

Core Area Metrics: Coefficient of Variation of Core Area Index (CAI_CV) – 2013
NLCD Class NLCD Class Value Range
11 Open Water 237 CAI_CV ≥ 0
21 Developed, Open Space 288 CAI_CV ≥ 0
22 Developed, Low Intensity 682 CAI_CV ≥ 0
23 Developed, Medium Intensity 561 CAI_CV ≥ 0
24 Developed, High Intensity 457 CAI_CV ≥ 0
31 Barren Land 212 CAI_CV ≥ 0
41 Deciduous Forest 332 CAI_CV ≥ 0
42 Evergreen Forest 165 CAI_CV ≥ 0
52 Shrub/Scrub 239 CAI_CV ≥ 0
71 Herbaceous 224 CAI_CV ≥ 0
81 Hay/Pasture 155 CAI_CV ≥ 0
82 Cultivated Crops 91 CAI_CV ≥ 0
90 Woody Wetlands 235 CAI_CV ≥ 0
95 Emergent Herbaceous Wetlands 264 CAI_CV ≥ 0

Core Area Metrics: Coefficient of Variation of Core Area Index (CAI_CV) – 2016
NLCD Class NLCD Class Value Range
11 Open Water 236 CAI_CV ≥ 0
79

21 Developed, Open Space 278 CAI_CV ≥ 0
22 Developed, Low Intensity 713 CAI_CV ≥ 0
23 Developed, Medium Intensity 620 CAI_CV ≥ 0
24 Developed, High Intensity 418 CAI_CV ≥ 0
31 Barren Land 300 CAI_CV ≥ 0
41 Deciduous Forest 316 CAI_CV ≥ 0
42 Evergreen Forest 165 CAI_CV ≥ 0
52 Shrub/Scrub 237 CAI_CV ≥ 0
71 Herbaceous 229 CAI_CV ≥ 0
81 Hay/Pasture 151 CAI_CV ≥ 0
82 Cultivated Crops 92 CAI_CV ≥ 0
90 Woody Wetlands 242 CAI_CV ≥ 0
95 Emergent Herbaceous Wetlands 264 CAI_CV ≥ 0

Core Area Metrics: Standard Deviation of Core Area Index (CAI_SD) – 2001
NLCD Class NLCD Class Value Range
11 Open Water 18.40 CAI_SD ≥ 0
21 Developed, Open Space 4.95 CAI_SD ≥ 0
22 Developed, Low Intensity 2.10 CAI_SD ≥ 0
23 Developed, Medium Intensity 2.16 CAI_SD ≥ 0
24 Developed, High Intensity 4.24 CAI_SD ≥ 0
31 Barren Land 10.70 CAI_SD ≥ 0
41 Deciduous Forest 4.17 CAI_SD ≥ 0
42 Evergreen Forest 13.90 CAI_SD ≥ 0
52 Shrub/Scrub 9.98 CAI_SD ≥ 0
71 Herbaceous 9.47 CAI_SD ≥ 0
81 Hay/Pasture 19.40 CAI_SD ≥ 0
82 Cultivated Crops 33.40 CAI_SD ≥ 0
90 Woody Wetlands 10.20 CAI_SD ≥ 0
95 Emergent Herbaceous Wetlands 10.10 CAI_SD ≥ 0

Core Area Metrics: Standard Deviation of Core Area Index (CAI_SD) – 2004
NLCD Class NLCD Class Value Range
11 Open Water 16.30 CAI_SD ≥ 0
21 Developed, Open Space 4.95 CAI_SD ≥ 0
22 Developed, Low Intensity 2.10 CAI_SD ≥ 0
23 Developed, Medium Intensity 2.16 CAI_SD ≥ 0
24 Developed, High Intensity 4.24 CAI_SD ≥ 0
80

31 Barren Land 10.60 CAI_SD ≥ 0
41 Deciduous Forest 4.13 CAI_SD ≥ 0
42 Evergreen Forest 13.90 CAI_SD ≥ 0
52 Shrub/Scrub 10.00 CAI_SD ≥ 0
71 Herbaceous 9.46 CAI_SD ≥ 0
81 Hay/Pasture 19.40 CAI_SD ≥ 0
82 Cultivated Crops 33.80 CAI_SD ≥ 0
90 Woody Wetlands 10.80 CAI_SD ≥ 0
95 Emergent Herbaceous Wetlands 9.42 CAI_SD ≥ 0

Core Area Metrics: Standard Deviation of Core Area Index (CAI_SD) – 2006
NLCD Class NLCD Class Value Range
11 Open Water 17.00 CAI_SD ≥ 0
21 Developed, Open Space 5.66 CAI_SD ≥ 0
22 Developed, Low Intensity 2.11 CAI_SD ≥ 0
23 Developed, Medium Intensity 2.06 CAI_SD ≥ 0
24 Developed, High Intensity 4.59 CAI_SD ≥ 0
31 Barren Land 10.90 CAI_SD ≥ 0
41 Deciduous Forest 4.13 CAI_SD ≥ 0
42 Evergreen Forest 13.80 CAI_SD ≥ 0
52 Shrub/Scrub 10.20 CAI_SD ≥ 0
71 Herbaceous 9.32 CAI_SD ≥ 0
81 Hay/Pasture 20.50 CAI_SD ≥ 0
82 Cultivated Crops 33.80 CAI_SD ≥ 0
90 Woody Wetlands 9.80 CAI_SD ≥ 0
95 Emergent Herbaceous Wetlands 8.78 CAI_SD ≥ 0

Core Area Metrics: Standard Deviation of Core Area Index (CAI_SD) – 2008
NLCD Class NLCD Class Value Range
11 Open Water 16.10 CAI_SD ≥ 0
21 Developed, Open Space 5.66 CAI_SD ≥ 0
22 Developed, Low Intensity 2.11 CAI_SD ≥ 0
23 Developed, Medium Intensity 2.06 CAI_SD ≥ 0
24 Developed, High Intensity 4.59 CAI_SD ≥ 0
31 Barren Land 10.70 CAI_SD ≥ 0
41 Deciduous Forest 4.09 CAI_SD ≥ 0
42 Evergreen Forest 13.90 CAI_SD ≥ 0
52 Shrub/Scrub 10.30 CAI_SD ≥ 0
81

71 Herbaceous 9.36 CAI_SD ≥ 0
81 Hay/Pasture 18.70 CAI_SD ≥ 0
82 Cultivated Crops 34.00 CAI_SD ≥ 0
90 Woody Wetlands 9.95 CAI_SD ≥ 0
95 Emergent Herbaceous Wetlands 9.34 CAI_SD ≥ 0

Core Area Metrics: Standard Deviation of Core Area Index (CAI_SD) – 2011
NLCD Class NLCD Class Value Range
11 Open Water 16.20 CAI_SD ≥ 0
21 Developed, Open Space 6.02 CAI_SD ≥ 0
22 Developed, Low Intensity 2.11 CAI_SD ≥ 0
23 Developed, Medium Intensity 2.25 CAI_SD ≥ 0
24 Developed, High Intensity 4.37 CAI_SD ≥ 0
31 Barren Land 11.10 CAI_SD ≥ 0
41 Deciduous Forest 3.94 CAI_SD ≥ 0
42 Evergreen Forest 13.50 CAI_SD ≥ 0
52 Shrub/Scrub 10.80 CAI_SD ≥ 0
71 Herbaceous 9.24 CAI_SD ≥ 0
81 Hay/Pasture 19.00 CAI_SD ≥ 0
82 Cultivated Crops 34.00 CAI_SD ≥ 0
90 Woody Wetlands 9.96 CAI_SD ≥ 0
95 Emergent Herbaceous Wetlands 9.53 CAI_SD ≥ 0

Core Area Metrics: Standard Deviation of Core Area Index (CAI_SD) – 2013
NLCD Class NLCD Class Value Range
11 Open Water 16.60 CAI_SD ≥ 0
21 Developed, Open Space 6.02 CAI_SD ≥ 0
22 Developed, Low Intensity 2.11 CAI_SD ≥ 0
23 Developed, Medium Intensity 2.25 CAI_SD ≥ 0
24 Developed, High Intensity 4.37 CAI_SD ≥ 0
31 Barren Land 11.30 CAI_SD ≥ 0
41 Deciduous Forest 4.17 CAI_SD ≥ 0
42 Evergreen Forest 13.50 CAI_SD ≥ 0
52 Shrub/Scrub 10.60 CAI_SD ≥ 0
71 Herbaceous 9.35 CAI_SD ≥ 0
81 Hay/Pasture 18.90 CAI_SD ≥ 0
82 Cultivated Crops 34.10 CAI_SD ≥ 0
90 Woody Wetlands 10.40 CAI_SD ≥ 0
82

95 Emergent Herbaceous Wetlands 9.28 CAI_SD ≥ 0

Core Area Metrics: Standard Deviation of Core Area Index (CAI_SD) – 2016
NLCD Class NLCD Class Value Range
11 Open Water 16.50 CAI_SD ≥ 0
21 Developed, Open Space 6.16 CAI_SD ≥ 0
22 Developed, Low Intensity 2.05 CAI_SD ≥ 0
23 Developed, Medium Intensity 2.44 CAI_SD ≥ 0
24 Developed, High Intensity 4.93 CAI_SD ≥ 0
31 Barren Land 11.50 CAI_SD ≥ 0
41 Deciduous Forest 5.48 CAI_SD ≥ 0
42 Evergreen Forest 13.40 CAI_SD ≥ 0
52 Shrub/Scrub 10.40 CAI_SD ≥ 0
71 Herbaceous 9.36 CAI_SD ≥ 0
81 Hay/Pasture 19.60 CAI_SD ≥ 0
82 Cultivated Crops 33.60 CAI_SD ≥ 0
90 Woody Wetlands 9.77 CAI_SD ≥ 0
95 Emergent Herbaceous Wetlands 9.13 CAI_SD ≥ 0

Core Area Metrics: Coefficient of Variation of Core Area (CORE_CV) – 2001
NLCD Class NLCD Class Value Range
11 Open Water 1,226 CORE_CV ≥ 0
21 Developed, Open Space 2,137 CORE_CV ≥ 0
22 Developed, Low Intensity 3,933 CORE_CV ≥ 0
23 Developed, Medium Intensity 1,926 CORE_CV ≥ 0
24 Developed, High Intensity 1,594 CORE_CV ≥ 0
31 Barren Land 2,357 CORE_CV ≥ 0
41 Deciduous Forest 351 CORE_CV ≥ 0
42 Evergreen Forest 3,979 CORE_CV ≥ 0
52 Shrub/Scrub 15,229 CORE_CV ≥ 0
71 Herbaceous 6,069 CORE_CV ≥ 0
81 Hay/Pasture 313 CORE_CV ≥ 0
82 Cultivated Crops 374 CORE_CV ≥ 0
90 Woody Wetlands 2,690 CORE_CV ≥ 0
95 Emergent Herbaceous Wetlands 1,840 CORE_CV ≥ 0

Core Area Metrics: Coefficient of Variation of Core Area (CORE_CV) – 2004
NLCD Class NLCD Class Value Range
83

11 Open Water 1,149 CORE_CV ≥ 0
21 Developed, Open Space 2,137 CORE_CV ≥ 0
22 Developed, Low Intensity 3,933 CORE_CV ≥ 0
23 Developed, Medium Intensity 1,926 CORE_CV ≥ 0
24 Developed, High Intensity 1,594 CORE_CV ≥ 0
31 Barren Land 1,914 CORE_CV ≥ 0
41 Deciduous Forest 355 CORE_CV ≥ 0
42 Evergreen Forest 3,983 CORE_CV ≥ 0
52 Shrub/Scrub 15,167 CORE_CV ≥ 0
71 Herbaceous 6,211 CORE_CV ≥ 0
81 Hay/Pasture 308 CORE_CV ≥ 0
82 Cultivated Crops 371 CORE_CV ≥ 0
90 Woody Wetlands 2,681 CORE_CV ≥ 0
95 Emergent Herbaceous Wetlands 2,049 CORE_CV ≥ 0

Core Area Metrics: Coefficient of Variation of Core Area (CORE_CV) – 2006
NLCD Class NLCD Class Value Range
11 Open Water 984 CORE_CV ≥ 0
21 Developed, Open Space 1,910 CORE_CV ≥ 0
22 Developed, Low Intensity 4,311 CORE_CV ≥ 0
23 Developed, Medium Intensity 1,949 CORE_CV ≥ 0
24 Developed, High Intensity 1,486 CORE_CV ≥ 0
31 Barren Land 1,697 CORE_CV ≥ 0
41 Deciduous Forest 355 CORE_CV ≥ 0
42 Evergreen Forest 3,974 CORE_CV ≥ 0
52 Shrub/Scrub 15,507 CORE_CV ≥ 0
71 Herbaceous 8,092 CORE_CV ≥ 0
81 Hay/Pasture 300 CORE_CV ≥ 0
82 Cultivated Crops 371 CORE_CV ≥ 0
90 Woody Wetlands 2,896 CORE_CV ≥ 0
95 Emergent Herbaceous Wetlands 2,102 CORE_CV ≥ 0

Core Area Metrics: Coefficient of Variation of Core Area (CORE_CV) – 2008
NLCD Class NLCD Class Value Range
11 Open Water 1,196 CORE_CV ≥ 0
21 Developed, Open Space 1,910 CORE_CV ≥ 0
22 Developed, Low Intensity 4,311 CORE_CV ≥ 0
23 Developed, Medium Intensity 1,949 CORE_CV ≥ 0
84

24 Developed, High Intensity 1,486 CORE_CV ≥ 0
31 Barren Land 1,900 CORE_CV ≥ 0
41 Deciduous Forest 359 CORE_CV ≥ 0
42 Evergreen Forest 3,971 CORE_CV ≥ 0
52 Shrub/Scrub 15,588 CORE_CV ≥ 0
71 Herbaceous 6,060 CORE_CV ≥ 0
81 Hay/Pasture 335 CORE_CV ≥ 0
82 Cultivated Crops 365 CORE_CV ≥ 0
90 Woody Wetlands 2,850 CORE_CV ≥ 0
95 Emergent Herbaceous Wetlands 2,002 CORE_CV ≥ 0

Core Area Metrics: Coefficient of Variation of Core Area (CORE_CV) – 2011
NLCD Class NLCD Class Value Range
11 Open Water 825 CORE_CV ≥ 0
21 Developed, Open Space 1,778 CORE_CV ≥ 0
22 Developed, Low Intensity 4,668 CORE_CV ≥ 0
23 Developed, Medium Intensity 1,821 CORE_CV ≥ 0
24 Developed, High Intensity 1,551 CORE_CV ≥ 0
31 Barren Land 1,383 CORE_CV ≥ 0
41 Deciduous Forest 375 CORE_CV ≥ 0
42 Evergreen Forest 4,378 CORE_CV ≥ 0
52 Shrub/Scrub 14,718 CORE_CV ≥ 0
71 Herbaceous 7,609 CORE_CV ≥ 0
81 Hay/Pasture 329 CORE_CV ≥ 0
82 Cultivated Crops 362 CORE_CV ≥ 0
90 Woody Wetlands 2,791 CORE_CV ≥ 0
95 Emergent Herbaceous Wetlands 2,008 CORE_CV ≥ 0

Core Area Metrics: Coefficient of Variation of Core Area (CORE_CV) – 2013
NLCD Class NLCD Class Value Range
11 Open Water 1,102 CORE_CV ≥ 0
21 Developed, Open Space 1,778 CORE_CV ≥ 0
22 Developed, Low Intensity 4,668 CORE_CV ≥ 0
23 Developed, Medium Intensity 1,821 CORE_CV ≥ 0
24 Developed, High Intensity 1,551 CORE_CV ≥ 0
31 Barren Land 1,452 CORE_CV ≥ 0
41 Deciduous Forest 351 CORE_CV ≥ 0
42 Evergreen Forest 4,387 CORE_CV ≥ 0
85

52 Shrub/Scrub 16,239 CORE_CV ≥ 0
71 Herbaceous 7,195 CORE_CV ≥ 0
81 Hay/Pasture 351 CORE_CV ≥ 0
82 Cultivated Crops 378 CORE_CV ≥ 0
90 Woody Wetlands 2,703 CORE_CV ≥ 0
95 Emergent Herbaceous Wetlands 2,089 CORE_CV ≥ 0

Core Area Metrics: Coefficient of Variation of Core Area (CORE_CV) – 2016
NLCD Class NLCD Class Value Range
11 Open Water 1,121 CORE_CV ≥ 0
21 Developed, Open Space 1,695 CORE_CV ≥ 0
22 Developed, Low Intensity 5,045 CORE_CV ≥ 0
23 Developed, Medium Intensity 1,821 CORE_CV ≥ 0
24 Developed, High Intensity 1,460 CORE_CV ≥ 0
31 Barren Land 1,383 CORE_CV ≥ 0
41 Deciduous Forest 400 CORE_CV ≥ 0
42 Evergreen Forest 4,386 CORE_CV ≥ 0
52 Shrub/Scrub 16,154 CORE_CV ≥ 0
71 Herbaceous 7,363 CORE_CV ≥ 0
81 Hay/Pasture 325 CORE_CV ≥ 0
82 Cultivated Crops 401 CORE_CV ≥ 0
90 Woody Wetlands 2,814 CORE_CV ≥ 0
95 Emergent Herbaceous Wetlands 2,097 CORE_CV ≥ 0

Contagion and Interspersion Metrics: Clumpy Index (CLUMPY) – 2001
NLCD Class NLCD Class Value Range
11 Open Water 0.84 -1 ≤ CLUMPY ≤ 1
21 Developed, Open Space 0.48 -1 ≤ CLUMPY ≤ 1
22 Developed, Low Intensity 0.48 -1 ≤ CLUMPY ≤ 1
23 Developed, Medium Intensity 0.40 -1 ≤ CLUMPY ≤ 1
24 Developed, High Intensity 0.49 -1 ≤ CLUMPY ≤ 1
31 Barren Land 0.78 -1 ≤ CLUMPY ≤ 1
41 Deciduous Forest 0.41 -1 ≤ CLUMPY ≤ 1
42 Evergreen Forest 0.82 -1 ≤ CLUMPY ≤ 1
52 Shrub/Scrub 0.71 -1 ≤ CLUMPY ≤ 1
71 Herbaceous 0.70 -1 ≤ CLUMPY ≤ 1
81 Hay/Pasture 0.80 -1 ≤ CLUMPY ≤ 1
82 Cultivated Crops 0.93 -1 ≤ CLUMPY ≤ 1
86

90 Woody Wetlands 0.85 -1 ≤ CLUMPY ≤ 1
95 Emergent Herbaceous Wetlands 0.77 -1 ≤ CLUMPY ≤ 1

Contagion and Interspersion Metrics: Clumpy Index (CLUMPY) – 2004
NLCD Class NLCD Class Value Range
11 Open Water 0.86 -1 ≤ CLUMPY ≤ 1
21 Developed, Open Space 0.48 -1 ≤ CLUMPY ≤ 1
22 Developed, Low Intensity 0.48 -1 ≤ CLUMPY ≤ 1
23 Developed, Medium Intensity 0.40 -1 ≤ CLUMPY ≤ 1
24 Developed, High Intensity 0.49 -1 ≤ CLUMPY ≤ 1
31 Barren Land 0.76 -1 ≤ CLUMPY ≤ 1
41 Deciduous Forest 0.41 -1 ≤ CLUMPY ≤ 1
42 Evergreen Forest 0.82 -1 ≤ CLUMPY ≤ 1
52 Shrub/Scrub 0.71 -1 ≤ CLUMPY ≤ 1
71 Herbaceous 0.70 -1 ≤ CLUMPY ≤ 1
81 Hay/Pasture 0.80 -1 ≤ CLUMPY ≤ 1
82 Cultivated Crops 0.93 -1 ≤ CLUMPY ≤ 1
90 Woody Wetlands 0.85 -1 ≤ CLUMPY ≤ 1
95 Emergent Herbaceous Wetlands 0.76 -1 ≤ CLUMPY ≤ 1

Contagion and Interspersion Metrics: Clumpy Index (CLUMPY) – 2006
NLCD Class NLCD Class Value Range
11 Open Water 0.87 -1 ≤ CLUMPY ≤ 1
21 Developed, Open Space 0.49 -1 ≤ CLUMPY ≤ 1
22 Developed, Low Intensity 0.46 -1 ≤ CLUMPY ≤ 1
23 Developed, Medium Intensity 0.40 -1 ≤ CLUMPY ≤ 1
24 Developed, High Intensity 0.49 -1 ≤ CLUMPY ≤ 1
31 Barren Land 0.76 -1 ≤ CLUMPY ≤ 1
41 Deciduous Forest 0.41 -1 ≤ CLUMPY ≤ 1
42 Evergreen Forest 0.83 -1 ≤ CLUMPY ≤ 1
52 Shrub/Scrub 0.72 -1 ≤ CLUMPY ≤ 1
71 Herbaceous 0.71 -1 ≤ CLUMPY ≤ 1
81 Hay/Pasture 0.81 -1 ≤ CLUMPY ≤ 1
82 Cultivated Crops 0.93 -1 ≤ CLUMPY ≤ 1
90 Woody Wetlands 0.84 -1 ≤ CLUMPY ≤ 1
95 Emergent Herbaceous Wetlands 0.76 -1 ≤ CLUMPY ≤ 1

87

Contagion and Interspersion Metrics: Clumpy Index (CLUMPY) – 2008
NLCD Class NLCD Class Value Range
11 Open Water 0.86 -1 ≤ CLUMPY ≤ 1
21 Developed, Open Space 0.49 -1 ≤ CLUMPY ≤ 1
22 Developed, Low Intensity 0.46 -1 ≤ CLUMPY ≤ 1
23 Developed, Medium Intensity 0.40 -1 ≤ CLUMPY ≤ 1
24 Developed, High Intensity 0.49 -1 ≤ CLUMPY ≤ 1
31 Barren Land 0.76 -1 ≤ CLUMPY ≤ 1
41 Deciduous Forest 0.41 -1 ≤ CLUMPY ≤ 1
42 Evergreen Forest 0.83 -1 ≤ CLUMPY ≤ 1
52 Shrub/Scrub 0.73 -1 ≤ CLUMPY ≤ 1
71 Herbaceous 0.72 -1 ≤ CLUMPY ≤ 1
81 Hay/Pasture 0.80 -1 ≤ CLUMPY ≤ 1
82 Cultivated Crops 0.93 -1 ≤ CLUMPY ≤ 1
90 Woody Wetlands 0.84 -1 ≤ CLUMPY ≤ 1
95 Emergent Herbaceous Wetlands 0.77 -1 ≤ CLUMPY ≤ 1

Contagion and Interspersion Metrics: Clumpy Index (CLUMPY) – 2011
NLCD Class NLCD Class Value Range
11 Open Water 0.81 -1 ≤ CLUMPY ≤ 1
21 Developed, Open Space 0.49 -1 ≤ CLUMPY ≤ 1
22 Developed, Low Intensity 0.44 -1 ≤ CLUMPY ≤ 1
23 Developed, Medium Intensity 0.40 -1 ≤ CLUMPY ≤ 1
24 Developed, High Intensity 0.47 -1 ≤ CLUMPY ≤ 1
31 Barren Land 0.77 -1 ≤ CLUMPY ≤ 1
41 Deciduous Forest 0.40 -1 ≤ CLUMPY ≤ 1
42 Evergreen Forest 0.82 -1 ≤ CLUMPY ≤ 1
52 Shrub/Scrub 0.76 -1 ≤ CLUMPY ≤ 1
71 Herbaceous 0.77 -1 ≤ CLUMPY ≤ 1
81 Hay/Pasture 0.80 -1 ≤ CLUMPY ≤ 1
82 Cultivated Crops 0.93 -1 ≤ CLUMPY ≤ 1
90 Woody Wetlands 0.84 -1 ≤ CLUMPY ≤ 1
95 Emergent Herbaceous Wetlands 0.76 -1 ≤ CLUMPY ≤ 1

Contagion and Interspersion Metrics: Clumpy Index (CLUMPY) – 2013
NLCD Class NLCD Class Value Range
11 Open Water 0.85 -1 ≤ CLUMPY ≤ 1
21 Developed, Open Space 0.49 -1 ≤ CLUMPY ≤ 1
88

22 Developed, Low Intensity 0.44 -1 ≤ CLUMPY ≤ 1
23 Developed, Medium Intensity 0.40 -1 ≤ CLUMPY ≤ 1
24 Developed, High Intensity 0.47 -1 ≤ CLUMPY ≤ 1
31 Barren Land 0.78 -1 ≤ CLUMPY ≤ 1
41 Deciduous Forest 0.41 -1 ≤ CLUMPY ≤ 1
42 Evergreen Forest 0.82 -1 ≤ CLUMPY ≤ 1
52 Shrub/Scrub 0.75 -1 ≤ CLUMPY ≤ 1
71 Herbaceous 0.76 -1 ≤ CLUMPY ≤ 1
81 Hay/Pasture 0.80 -1 ≤ CLUMPY ≤ 1
82 Cultivated Crops 0.93 -1 ≤ CLUMPY ≤ 1
90 Woody Wetlands 0.85 -1 ≤ CLUMPY ≤ 1
95 Emergent Herbaceous Wetlands 0.75 -1 ≤ CLUMPY ≤ 1

Contagion and Interspersion Metrics: Clumpy Index (CLUMPY) – 2016
NLCD Class NLCD Class Value Range
11 Open Water 0.87 -1 ≤ CLUMPY ≤ 1
21 Developed, Open Space 0.50 -1 ≤ CLUMPY ≤ 1
22 Developed, Low Intensity 0.42 -1 ≤ CLUMPY ≤ 1
23 Developed, Medium Intensity 0.39 -1 ≤ CLUMPY ≤ 1
24 Developed, High Intensity 0.47 -1 ≤ CLUMPY ≤ 1
31 Barren Land 0.78 -1 ≤ CLUMPY ≤ 1
41 Deciduous Forest 0.44 -1 ≤ CLUMPY ≤ 1
42 Evergreen Forest 0.82 -1 ≤ CLUMPY ≤ 1
52 Shrub/Scrub 0.74 -1 ≤ CLUMPY ≤ 1
71 Herbaceous 0.73 -1 ≤ CLUMPY ≤ 1
81 Hay/Pasture 0.81 -1 ≤ CLUMPY ≤ 1
82 Cultivated Crops 0.93 -1 ≤ CLUMPY ≤ 1
90 Woody Wetlands 0.85 -1 ≤ CLUMPY ≤ 1
95 Emergent Herbaceous Wetlands 0.76 -1 ≤ CLUMPY ≤ 1
Diversity Metrics
Shannon's Diversity Index (SHDI) Range: SHDI ≥ 1
2001 2004 2006 2008 2011 2013 2016
0.733 0.730 0.756 0.765 0.828 0.821 0.807
Simpson's Diversity Index (SIDI) Range: 0 ≤ SIDI < 1
89

2001 2004 2006 2008 2011 2013 2016
0.357 0.355 0.369 0.376 0.427 0.418 0.400
Patch Richness (PR) Range: PR ≥ 1
2001 2004 2006 2008 2011 2013 2016
14 14 14 14 14 14 14
Patch Richness Density (PRD) Range: PRD ≥ 0
2001 2004 2006 2008 2011 2013 2016
0.00055 0.00055 0.00055 0.00055 0.00055 0.00055 0.00055
Shannon's evenness index (SHEI) Range: 0 ≤ SHEI < 1
2001 2004 2006 2008 2011 2013 2016
0.278 0.277 0.286 0.290 0.314 0.311 0.306
Simpson's evenness index (SIEI) Range: 0 < SIEI ≤ 1
2001 2004 2006 2008 2011 2013 2016
0.384 0.382 0.398 0.405 0.460 0.450 0.431

90

Appendix C LII values and LCM Values for Model Validation
Table of LII value and LCM values (LII_LCM variable in R Markdown)
OID LII_Value LCM_Value
1 0.507061 88
2 0.540633 100
3 0.512471 100
4 0.536162 70
5 0.438775 70
6 0.465446 100
7 0.465215 67
8 0.481052 100
9 0.508359 100
10 0.196712 28
11 0.523356 98
12 0.544645 70
13 0.494623 98
14 0.469222 92
15 0.511743 100
16 0.487586 98
17 0.322849 20
18 0.537398 100
19 0.485023 100
91

20 0.507781 100
21 0.517532 70
22 0.486459 88
23 0.488039 70
24 0.517731 100
25 0.512488 100
26 0.486303 63
27 0.474127 98
28 0.491344 95
29 0.515071 93
30 0.481362 65
31 0.490334 100
32 0.473876 26
33 0.544825 100
34 0.499425 98
35 0.532944 100
36 0.519079 70
37 0.538403 83
38 0.464194 60
39 0.487845 67
40 0.491651 45
41 0.466127 100
42 0.497011 83
92

43 0.447463 70
44 0.511479 100
45 0.461048 62
46 0.515468 56
47 0.513639 88
48 0.379457 100
49 0.521584 88
50 0.46471 37
51 0.510081 100
52 0.531166 70
53 0.586949 100
54 0.473013 70
55 0.47359 66
56 0.505251 70
57 0.482825 53
58 0.531216 34
59 0.48936 70
60 0.471661 100
61 0.479537 100
62 0.472096 70
63 0.517264 100
64 0.438283 100
65 0.485968 70
93

66 0.476776 24
67 0.498625 98
68 0.509831 88
69 0.418751 99
70 0.493656 100
71 0.519828 88
72 0.525642 30
73 0.536188 88
74 0.461057 91
75 0.182372 37
76 0.48967 100
77 0.508072 100
78 0.504031 100
79 0.459072 97
80 0.505294 100
81 0.486359 88
82 0.501651 98
83 0.493336 70
84 0.488122 70
85 0.504493 70
86 0.503706 94
87 0.447484 44
88 0.507609 100
94

89 0.504074 86
90 0.510223 100
91 0.509943 99
92 0.201698 0
93 0.48267 70
94 0.182823 36
95 0.382976 22
96 0.50254 70
97 0.503198 26
98 0.485859 70
99 0.521986 100
100 0.478407 70

95

Table of LII values in Protected Areas or Multiple-Use Areas (LII_PADUS_100 in R
Markdown)

OID RASTERVALU group
1 0.479084 Protected
2 0.480042 Protected
3 0.501907 Protected
4 0.501019 Protected
5 0.474674 Protected
6 0.547407 Protected
7 0.490589 Protected
8 0.497564 Protected
9 0.53569 Protected
10 0.478197 Protected
11 0.511125 Protected
12 0.500255 Protected
13 0.51147 Protected
14 0.520881 Protected
15 0.517646 Protected
16 0.488333 Protected
17 0.542949 Protected
18 0.559083 Protected
19 0.490912 Protected
20 0.453005 Protected
96

21 0.490856 Protected
22 0.490633 Protected
23 0.468514 Protected
24 0.476357 Protected
25 0.471646 Protected
26 0.478189 Protected
27 0.529337 Protected
28 0.463209 Protected
29 0.52526 Protected
30 0.523877 Protected
31 0.457668 Protected
32 0.50062 Protected
33 0.543463 Protected
34 0.498557 Protected
35 0.463255 Protected
36 0.541929 Protected
37 0.496821 Protected
38 0.557516 Protected
39 0.472581 Protected
40 0.478606 Protected
41 0.489 Protected
42 0.496746 Protected
43 0.521122 Protected
97

44 0.503573 Protected
45 0.4883 Protected
46 0.457678 Protected
47 0.529721 Protected
48 0.436923 Protected
49 0.475973 Protected
50 0.528192 Protected
51 0.478312 Protected
52 0.517202 Protected
53 0.495025 Protected
54 0.434089 Protected
55 0.549679 Protected
56 0.490334 Protected
57 0.550763 Protected
58 0.477253 Protected
59 0.455615 Protected
60 0.523612 Protected
61 0.519409 Protected
62 0.489472 Protected
63 0.483062 Protected
64 0.477214 Protected
65 0.500976 Protected
66 0.532278 Protected
98

67 0.50061 Protected
68 0.501019 Protected
69 0.480833 Protected
70 0.485737 Protected
71 0.483849 Protected
72 0.491054 Protected
73 0.479333 Protected
74 0.534523 Protected
75 0.540275 Protected
76 0.478808 Protected
77 0.522823 Protected
78 0.555473 Protected
79 0.547746 Protected
80 0.457162 Protected
81 0.483737 Protected
82 0.476213 Protected
83 0.517449 Protected
84 0.53136 Protected
85 0.443585 Protected
86 0.436516 Protected
87 0.494409 Protected
88 0.549679 Protected
89 0.526437 Protected
99

90 0.440661 Protected
91 0.518173 Protected
92 0.482767 Protected
93 0.524042 Protected
94 0.470283 Protected
95 0.463179 Protected
96 0.556959 Protected
97 0.494164 Protected
98 0.469658 Protected
99 0.481625 Protected
100 0.49594 Protected
1 0.468572 Multiple-Use
2 0.520155 Multiple-Use
3 0.47186 Multiple-Use
4 0.498978 Multiple-Use
5 0.511508 Multiple-Use
6 0.483726 Multiple-Use
7 0.487944 Multiple-Use
8 0.493899 Multiple-Use
9 0.476737 Multiple-Use
10 0.466534 Multiple-Use
11 0.460113 Multiple-Use
12 0.508826 Multiple-Use
100

13 0.493513 Multiple-Use
14 0.519825 Multiple-Use
15 0.511102 Multiple-Use
16 0.560985 Multiple-Use
17 0.531744 Multiple-Use
18 0.478779 Multiple-Use
19 0.493602 Multiple-Use
20 0.464205 Multiple-Use
21 0.451005 Multiple-Use
22 0.48675 Multiple-Use
23 0.478822 Multiple-Use
24 0.480982 Multiple-Use
25 0.489191 Multiple-Use
26 0.462792 Multiple-Use
27 0.461041 Multiple-Use
28 0.475372 Multiple-Use
29 0.489597 Multiple-Use
30 0.517958 Multiple-Use
31 0.49924 Multiple-Use
32 0.494106 Multiple-Use
33 0.504031 Multiple-Use
34 0.47124 Multiple-Use
35 0.540468 Multiple-Use
101

36 0.459279 Multiple-Use
37 0.452474 Multiple-Use
38 0.478274 Multiple-Use
39 0.483011 Multiple-Use
40 0.510597 Multiple-Use
41 0.471559 Multiple-Use
42 0.473022 Multiple-Use
43 0.417273 Multiple-Use
44 0.477178 Multiple-Use
45 0.436605 Multiple-Use
46 0.555461 Multiple-Use
47 0.464811 Multiple-Use
48 0.525161 Multiple-Use
49 0.472316 Multiple-Use
50 0.451213 Multiple-Use
51 0.477264 Multiple-Use
52 0.520422 Multiple-Use
53 0.50793 Multiple-Use
54 0.452859 Multiple-Use
55 0.461041 Multiple-Use
56 0.483855 Multiple-Use
57 0.498682 Multiple-Use
58 0.467832 Multiple-Use
102

59 0.488314 Multiple-Use
60 0.479803 Multiple-Use
61 0.492889 Multiple-Use
62 0.461902 Multiple-Use
63 0.478678 Multiple-Use
64 0.489302 Multiple-Use
65 0.507905 Multiple-Use
66 0.512495 Multiple-Use
67 0.551705 Multiple-Use
68 0.466424 Multiple-Use
69 0.488655 Multiple-Use
70 0.462115 Multiple-Use
71 0.493628 Multiple-Use
72 0.501725 Multiple-Use
73 0.455363 Multiple-Use
74 0.478705 Multiple-Use
75 0.6911 Multiple-Use
76 0.477769 Multiple-Use
77 0.51162 Multiple-Use
78 0.512147 Multiple-Use
79 0.452351 Multiple-Use
80 0.477264 Multiple-Use
81 0.508263 Multiple-Use
103

82 0.51204 Multiple-Use
83 0.498594 Multiple-Use
84 0.46593 Multiple-Use
85 0.508393 Multiple-Use
86 0.49344 Multiple-Use
87 0.477572 Multiple-Use
88 0.464205 Multiple-Use
89 0.480495 Multiple-Use
90 0.464833 Multiple-Use
91 0.509816 Multiple-Use
92 0.517193 Multiple-Use
93 0.504031 Multiple-Use
94 0.484374 Multiple-Use
95 0.461041 Multiple-Use
96 0.460261 Multiple-Use
97 0.454689 Multiple-Use
98 0.495072 Multiple-Use
99 0.491028 Multiple-Use
100 0.484863 Multiple-Use

Abstract (if available)

Abstract The Bureau of Land Management (BLM) is instrumental in connecting people with public lands by providing and protecting opportunities to enjoy and use our country’s resources. Understanding the cumulative effects of resource management programs is crucial for decision makers to develop effective land management practices and appropriate allocation of funding and resources. A comprehensive, standardized, and transparent GIS workflow can help visualize and analyze ecological integrity, landscape patterns and processes, and promote a consistent Cumulative Effects Analysis (CEA) and collaborative management across jurisdiction boundaries. ❧ This research evaluates the cumulative impacts of resource management programs in the BLM Carlsbad Field Office (CFO), New Mexico by incorporating ecological integrity indicators, resource- and stressor-based metrics, and landscape metrics to create a Landscape Integrity Index (LII). Two resource management programs, Vegetative Communities and Minerals – Leasables – Oil and Gas, were selected as the programs of interest for this study. The LII model considers the management goals and objectives in the Draft BLM CFO Resource Management Plan (RMP) to identify the necessary indicators and metrics. These indicators and metrics were each scored for their site impact, distance decay function, or landscape metrics through the use of a Composite Scoring System, and then combined into a single map. The resulting map with the LII values shows areas of low landscape integrity near the urban and agricultural areas in CFO planning area and high landscape integrity near central and southwest corner of CFO. CEA practitioners and land managers will be able to address management goals and objectives, conduct a more systematic and consistent analysis with relevant indicators and metrics, and visualize landscape integrity using the LII framework.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Exploring remote sensing and geographic information systems technologies to understand vegetation changes in response to land management practices at Finke Gorge National Park, Australia Between ...

PDF

Questioning the cause of calamity: using remotely sensed data to assess successive fire events

PDF

Using geospatial technology to establish marsh bird monitoring sites for a pilot study in Maine in accordance with the North American Marsh Bird Monitoring Protocol

PDF

Impacts of vegetation management on wildfire severity: a study of the 2021 Caldor fire

PDF

Use of remotely sensed imagery to map sudden oak death (Phytophthora ramorum) in the Santa Cruz Mountains

PDF

The movement of Mexican migration and its impact based on a GIS geospatial database

PDF

Using Maxent to model the distribution of prehistoric agricultural features in a portion of the Hōkūli‘a subdivision in Kona, Hawai‘i

Asset Metadata

Creator Lee, Liling (author)

Core Title Using Landscape Integrity Index to evaluate the cumulative impacts of BLM resource management programs

School College of Letters, Arts and Sciences

Degree Master of Science

Degree Program Geographic Information Science and Technology

Publication Date 02/07/2020

Defense Date 12/13/2019

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag BLM,Bureau of Land Management,CEA,cumulative effects analysis,cumulative impacts,ecological integrity indicators,Landscape Integrity Index,landscape metrics,OAI-PMH Harvest

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Bernstein, Jennifer Moore (committee chair), Lee, Su Jin (committee member), Wilson, John P. (committee member)

Creator Email liling2lee@gmail.com,lilingle@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-265496

Unique identifier UC11675265

Identifier etd-LeeLiling-8131.pdf (filename),usctheses-c89-265496 (legacy record id)

Legacy Identifier etd-LeeLiling-8131.pdf

Dmrecord 265496

Document Type Thesis

Rights Lee, Liling

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA