University of Southern California Dissertations and Theses

Building a geodatabase design for American Pika presence and absence data

(USC Thesis Other)

Building a geodatabase design for American Pika presence and absence data

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

Building a Geodatabase Design for American Pika Presence and Absence Data

by

Kyle Krueger Burke

A Thesis Presented to the
Faculty of the USC Graduate School
University of Southern California
In Partial Fulfillment of the
Requirements for the Degree
Master of Science
(Geographic Information Science and Technology)

May 2018

Copyright © 2018 by Kyle Burke

To My Wife:
For without her support this
would not have been possible

iv
Table of Contents
List of Figures ................................................................................................................................ vi
List of Tables ................................................................................................................................ vii
Acknowledgements ...................................................................................................................... viii
List of Abbreviations ..................................................................................................................... ix
Abstract ........................................................................................................................................... x
Chapter 1 Introduction .................................................................................................................... 1
1.1. Background .........................................................................................................................2
1.1.1. Study Area .................................................................................................................2
1.1.2. The American Pika (Ochotona princeps) ...................................................................3
1.1.3. Field Collection Techniques ......................................................................................4
1.2. Motivation ...........................................................................................................................5
1.3. Objectives ...........................................................................................................................6
1.4. Overview of Project Methods .............................................................................................6
1.4.1. Geodatabase Design ...................................................................................................7
1.4.2. Data Cleaning.............................................................................................................7
1.4.3. Digitization of Talus ..................................................................................................7
1.4.4. Importation and Testing of Data ................................................................................7
1.5. Structure of this Paper .........................................................................................................8
Chapter 2 Related Work.................................................................................................................. 9
2.1. Geodatabase Design ............................................................................................................9
2.2. Pika Habitat and Climate Change .....................................................................................11
Chapter 3 Methods ........................................................................................................................ 15
3.1 Methods Discussion ...........................................................................................................15
3.1.1. Geodatabase Design .................................................................................................15
v
3.1.2. Data Overview .........................................................................................................18
3.1.3. Detail of Relationships .............................................................................................19
3.2. Data, Software, and Hardware Requirements ...................................................................20
3.2.1. Data Requirements ...................................................................................................20
3.2.2. Software Requirements ............................................................................................21
3.2.3. Hardware Requirements...........................................................................................21
3.3. Timeline ............................................................................................................................22
3.4. Detailed Methods ..............................................................................................................23
3.4.1. Data Standardization and Importing ........................................................................23
3.4.2. Digitizing Polygons .................................................................................................28
3.4.3. Queries .....................................................................................................................30
Chapter 4 Results .......................................................................................................................... 31
4.1. Final Design and Data .......................................................................................................31
4.1.1. Final Design/Schema ...............................................................................................31
4.1.2. Geodatabase Data and Records ................................................................................31
4.2. Query Results ....................................................................................................................34
4.3. Summarizing of Results and Errors ..................................................................................43
Chapter 5 Conclusion .................................................................................................................... 46
5.1. Project Lessons .................................................................................................................46
5.2. Current State of Geodatabase ............................................................................................47
5.3. Future Work ......................................................................................................................48
References ..................................................................................................................................... 51
Appendices .................................................................................................................................... 54
Appendix A – Initial Entity Relationship Diagram ...................................................................... 54

vi
List of Figures
Figure 1 Original Map from Dr. Beever’s Research of the Great Basin Region with locations of
pika (Beever et al. 2011) ......................................................................................................... 3
Figure 2 Final ERD with changes to Feature Classes and Attributes ........................................... 17
Figure 3 Feature Dataset, Feature Classes, and Joins ................................................................... 19
Figure 4 Depictions of Relationship Classes in ArcMap .............................................................. 20
Figure 5 Example of Digitized talus on www.caltopo.com .......................................................... 21
Figure 6 Unformatted Detection Data (Latitude and Longitude Redacted) ................................. 24
Figure 7 Detection Feature Class Formatted Fields ...................................................................... 25
Figure 8 Example of Attributes for Patch ..................................................................................... 27
Figure 9 Example of Digitization ................................................................................................. 29
Figure 10 Visualization of all data showing Pika habitat in the study area of the Great Basin
Region (Patch/Polygon data is not visible at this scale) ........................................................ 33
Figure 11 Query and Results showing locations of presence/absence points taken during the 2016
field season ............................................................................................................................ 35
Figure 12 Queries for finding the maximum and minimum elevation of pika detections ............ 36
Figure 13 Results for finding the maximum and minimum elevation of pika detections............. 37
Figure 14 Query and Results of Unsurveyed Potential Patches ................................................... 38
Figure 15 Query and Results of iButton Locations with Pika Occupancy in both 2010 and 2017
............................................................................................................................................... 39
Figure 16 Query and Results of Highest Temperature Recorded ................................................. 40
Figure 17 Query and Results of Sites that Lost Pika Population between the 1990’s field seasons
to the 2003-2008 field seasons .............................................................................................. 41
Figure 18 Results for Query 6 ....................................................................................................... 42
Figure 19 Relationship of nested features example showing the Arc Dome Site with ARDO1e
Talus Patch related to it with the Presence/Absence Point 1442 related to both .................. 43
Figure 20 Example of Site with all Feature Classes ..................................................................... 45
vii
List of Tables
Table 1 Project Timeline ............................................................................................................... 23
Table 2 Queries for Testing Geodatabase Functionality ............................................................... 30
Table 3 Final Count of Imported Records .................................................................................... 32

viii
Acknowledgements

I would especially like to say thank you to Dr. Eric A. Beever whose research and enthusiastic
support for this project I owe everything. Special thanks to my father for the seemingly endless
discussions which eventually led to this paper and to my wife and son for allowing me hours of
seclusion when necessary. Thank you to my USC thesis committee for the guidance and support
to complete this project.

ix
List of Abbreviations
ERD Entity Relationship Diagram
GIS Geographic information system
GISci Geographic information science
SSI Spatial Sciences Institute
USC University of Southern California
USGS United State Geological Survey
x
Abstract
A United States Geological Survey (USGS) researcher has been studying impacts of
climate change on American Pika (Ochotona princeps) from the mid 1990’s through 2017. This
project aims to contribute to research on the American Pika by building a geodatabase to store
and provide access to data on pika populations throughout the Great Basin region of the Western
United States. The geodatabase contains pika presence and absence data for locations of talus,
which includes habitat areas that have been previously surveyed or may be potentially surveyed
in the future. The project used formatted data provided from field surveyed talus that have been
digitized on www.caltopo.com, digitized new talus that have been more recently surveyed, and
imported GPS points for presence/absence captured in Excel spreadsheets.
The end result of this project was a geodatabase that housed presence/absence points,
talus polygons, site locations, temperature sensor locations, and temperature/relative humidity
data. Several queries were completed that show proper importation and relationships of all data.
Working closely with project researchers, this study allows for database expansion as needed for
future research needs.
Studying presence/absence of American Pika allows for further understanding of climatic
impacts in niche habitats that are especially susceptible to environmental change. This project
also provides the opportunity for improved analysis and long term data storage relating to these
presence/absence locations throughout the Great Basin region. The end result supports expansion
of the database structure for future field seasons and data inclusion.

1

Chapter 1 Introduction
This project aims to create a geodatabase of a longitudinal study on American Pika (Ochotona
princeps) habitat areas from previously surveyed and potential survey locations in the Great
Basin region of the Western United States. The end result aims to support speedy analysis of
habitat changes of small montane animals that have been studied for more than 20 years. Data
housed in this geodatabase is specifically related to the presence and absence of American Pika
at surveyed talus (rocky slopes habitat) locations and potential future survey talus. These data
have been previously compiled and, therefore, this effort augments work performed by USGS
scientist, Dr. Erik Beever, by creating a single database to contain several years of data collected
in the field. The creation of a database to house the field collection data also allows for additions
of future field seasons. Beever’s project work focuses on the climate change impact on pika and
how they may be used as an indicator species for future climate impacts (Beever et al. 2003;
Beever et al. 2010; Beever et al. 2011; Beever et al. 2013). The American Pika has a very limited
range of habitat in which it has the ability to survive. Because of that niche habitat, pika have
developed specific adaptations to allow for survival. Dr. Beever’s research aims to investigate
how changes in climate impact on a species like the American Pika and what that might tell us
about how other species and environments react longer term.
The creation of a geodatabase compiled collected data and added the ability to swiftly
import future data from every field season. The goal is to allow for swifter analysis of these
collected data, as well as storage of all data in one singular location. The geodatabase allows for
faster exploration and analysis of collected data to see any gaps that may exist for upcoming field
seasons or modeling. The ability to continuously build upon previously collected work is an
advantage to researchers who may want to incorporate these data into future work. While the
2

ability to swiftly and accurately share data is a great advantage, the main goal of this study is to
allow researchers to have spatially referenced talus with the associated presence/absence point
data and attributes all housed in one location.
Section 1.1 provides further background on the project research. Section 1.2 provides
details of the motivation behind this project. Section 1.3 discusses the objectives that this project
looks to accomplish. Section 1.4 looks at an overview of the project methods and Section 1.5
breaks down the structure of the thesis.
1.1. Background
The Background section is broken into three subsections. 1.1.1 discusses the background
on the research area, 1.1.2 discusses the American Pika (Ochotona princeps) biology and
ecology and, 1.1.3 discusses the field collection techniques.
1.1.1. Study Area
The study area for this project focuses primarily on the Great Basin hydrographic region
located in the western United States. The Great Basin includes an area of roughly 500,000km²
that stretches between the Sierra Nevada and Rocky Mountains (see Figure 1). The general
geography of the region consists of north-south mountain ranges with a cold-temperate
semidesert climate. Within the 20
th
century, the Great Basin has warmed an average of 0.3 to
0.6°C and there are expectations that it will rise a further 2.5 to 4°C by 2100 (Rowe and Terry
2014). The change in temperature along with the great biodiversity of the region, especially
among small mammals, make it an exceptional study area in regards to climatic impact.
3

1.1.2. The American Pika (Ochotona princeps)
American Pika (Ochotona princeps) (hereafter referred to as pika) are small mammals
that have limited, discontinuous, and isolated range on the high rocky slopes of the western
Figure 1. Original Map from Dr. Beever’s Research of the Great Basin Region with locations
of pika (Beever et al. 2011).
4

North American mountains (Millar and Westfall 2010). Pika require a generally cool climate
with higher than average snowfall during winter to insulate them from extreme cold, as they do
not hibernate (Beever et al. 2010). While typically found at elevations anywhere from sea level
to 3000m along the northern edge of their range, at the southern edge most are found at 2500m
and higher due to their cool weather climate requirement. These high montane requirements
have create niche or island habitats that isolate population groups in specific rocky areas
surrounded by suitable vegetation (Smith and Weston 1990).
With mean temperatures in North America expected to rise >2°C by the middle of the
21
st
century and >4°C by the late 21
st
century, the impact of climate change will be felt across the
continent. Further, North America is expected to have daily extremes in excess of 5°C warmer
than current temperatures and >80% of future years are predicted to have snow falls less than the
middle of the 20
th
century levels (IPCC 2015). These factors along with the understanding that
alpine areas will be significantly impacted because of climate change (Villers-Ruiz and
Castaneda-Aguado 2013) make the pika population and range an important part of identifying
how temperature increase will impact montane species.
Another major drawback for pika adaptability is their relatively small dispersal area.
Found to be 0.8-3.0km (maximum and under cool conditions), the limited range of dispersal
poses severe restrictions on the pika’s ability to diffuse to new habitat should the current become
inhospitable (Beever et al. 2003; Hafner 1993; Rodhouse et al 2010).
1.1.3. Field Collection Techniques
The field collection techniques used to collect the presence/absence data are detailed
below. During multiple fields seasons talus locations that where known to have pika populations
where surveyed for continuing populations. Each location was surveyed for 8 hours along with
5

all talus locations within 3km as designated by the maximum known dispersal range. Field
surveys included walking 50 m long transect lines across talus slopes around 15 vertical meters
apart, using handheld GPS units to record the location of any pika sign, such has haypiles, feces,
sighting, etc.
During these surveys Celsius degree and relative humidity sensor readings were recorded.
These readings were obtained by placing temperature sensors roughly 80 cm under the surface of
the talus at ideal living areas for pika. The sensors used were DS1921G Thermochron i-Buttons
manufactured by Maxim Integrated Products of Sunnyvale, California, USA. The sensors were
placed at locations throughout the Great Basin research region and recorded temperature to the
nearest 0.1°C every two hours for the first five months after placement and every four hours after
the initial period of time (Beever et al. 2010). During original field seasons all sensors collected
data using the previously mentioned collection techniques. This has changed in more recent field
seasons to only collect a reading every four hours. All sensors continue in operation until they no
longer function or are replaced. The sensors are often replaced due to loss of function or inability
to find the old sensor. The data that have been collected in previous years, and will continue to
be collected in future years, are the main focus of this study and were used to create the
geodatabase.
1.2. Motivation
According to the Intergovernmental Panel on Climate Change (IPCC) the results of
numerous studies have led to the general consensus that humans have caused a rapid change in
climate (IPCC 2015). In the coming years, the ability to further research and model the impacts
of climate change will become more important. Researching specific species allows for scientists
to attempt to understand the future impacts of climate change on the world. Pika are one such
6

species that can be used as a predictor of climate impact because of its susceptibility to dramatic
shifts in temperature. The overall goal of Dr. Beever’s study is to further the understanding of
how climate impacts the biology and ecology of specific areas such as montane or coastal zones.
In Dr. Beever’s project, the focus is montane zones. This project will allow researchers to further
analyze and study climate change by building a geodatabase that allows for queries and spatial
analysis to provide insight into these goals.
1.3. Objectives
In this study there are two primary objectives. The first objective is to build a
geodatabase to house the majority of Dr. Beever’s collected research which allows for the ability
to quickly import, analyze, and present results of the presence/absence analysis. The completed
geodatabase permits numerous years of data to be housed together and will create a system of
simple data importation for future data collections efforts. This eases the current data storage
issues surrounding multiple Excel spreadsheets. The geodatabase also allows for almost instant
importation of field data into a permanent structure. The second objective is to build a
framework for a geodatabase which other researchers will be able to use as a template for future
research projects of a similar nature.
1.4. Overview of Project Methods
The following is a brief overview of the methods for this project and process involved.
This segment is broken into multiple sections starting with 1.3.1 which discusses the geodatabase
design. Section 1.3.2 describes the cleanup of data that was required for importation. Section
1.3.3 describes the digitization of talus, which was needed to expand the previous data. Finally,
1.3.4 discusses the data importation and testing of the geodatabase.
7

1.4.1. Geodatabase Design
The first step taken in this project was the design of the geodatabase. This was completed
by first designing an entity relationship diagram (ERD) and getting feedback on the initial
structure. Revisions and changes were then made based on that feedback. The geodatabase
structure was then built in ArcCatalog 10.5 and importation of an initial sample dataset was
completed to test functionality.
1.4.2. Data Cleaning
The second step that was taken in this project was cleaning up data provided. Data were
previously housed in numerous Excel spreadsheets that resulted in several different versions and
years of data. These spreadsheets were formatted in numerous different ways by several different
people and needed a lot of effort to assign proper naming conventions for all attributes. Once
completed this greatly eased importation into the geodatabase.
1.4.3. Digitization of Talus
The next step for this project involved digitizing the remaining talus via
www.Caltopo.com. This part of the project was previously started by Dr. Beever and other
researchers in an attempt to begin to assign point data to specific talus/polygons. Although
consideration was given to continuing to use www.Caltopo.com in an effort to be consistent with
previous research due to attributes not being assigned properly, it was necessary to complete this
effort in ArcMap 10.5 using an ArcGIS Online base layer.
1.4.4. Importation and Testing of Data
The final step was a combination of importing all data into the completed geodatabase
and then testing functionality. The data importation was much faster during this step due to the
8

numerous measures taken in early areas of the project to ensure proper formatting. Queries were
also written based on several research objectives discussed with Dr. Beever (i.e. number of
continuously extant sites between specific years). Once importation was completed, the queries
were run which demonstrated proper database functionality.
1.5. Structure of this Paper
The next chapter discusses research related to this project providing background on
geodatabase design, pika ecology, and climate change studies using wildlife. The third chapter
outlines the methods used to complete this project. The fourth and fifth chapters discuss the
results and conclusions respectively. Following the fifth chapter is a list of references.

9

Chapter 2 Related Work
This chapter looks at the research and background for geodatabase design, the ecology and
biology of the American Pika (Ochotona princeps), the impact of climate on their habitat, and
the reasons for studying this particular species. Section 2.1 discusses the research for building
and populating wildlife biology data in a geodatabase. Section 2.2 explores the biology, ecology,
and climate change studies of pika.
2.1. Geodatabase Design
For this project, using examples of geodatabase design for wildlife tracking (Urbano and
Cagnacci 2014), as well as multiple sources on geodatabase design for archaeological sites, was
invaluable (Mocanu and Velicanu 2011; Gonzalez-Tennant 2009; Breunig et al. 2016). While
using the geodatabase design for wildlife tracking may make perfect sense, the use of
archaeological geodatabase design may be less obvious. For example, with the island nature of
habitable sites for pika (i.e. talus) the model of an archaeological site geodatabase becomes more
suitable. Both this geodatabase and archaeological geodatabases are usually bounded by a
specific area. Archaeological geodatabases use the excavation site, this project used small talus
areas. Outside of these areas, limited factors tend to be included in the database giving both
designs island like results. Both archaeological and talus sites are required to have point data
within the polygon specifying specific locations which data and attributes are recorded for each.
As an example, for archaeological sites, excavators, artifacts, provenance, etc. and for talus, data
collectors, pika sign, temperature, etc. These correlations allow for design techniques to be used
from archaeological geodatabases and applied to this project.
10

This project also aims to fill a void within current research for creating geodatabases for
niche species. Wildlife tracking databases, such as Urbano and Cagnacci (2014) use thousands of
individual data points and then are able to extrapolate a range for the tracked species. This
geodatabase model differs and was developed for species that have a specific, known set of
criteria in which they persist (i.e. niche species). Instead of tracking animal movement and
analyzing range and area based on movement, this database houses survey data for specific sites
know to contain the attributes for pika to survive.
Another key consideration with any wildlife tracking is there tends to be very large
datasets that need to be “securely, consistently and efficiently managed” as to allow for any
number of changes and people to use the dataset (Urbano and Cagnacci 2014). Wildlife spatial
data also tends to be housed locally and does not set any standards for interoperability which can
then require several separate procedures when analyzing the same datasets (Urbano et al. 2010).
These considerations were taken into account when deciding to develop a geodatabase to house
all previous information together for faster importation/exportation and analysis along with the
idea that because of the cost associated with surveying and collecting wildlife data it is of great
importance to share this research with many other researchers (Urbano and Cagnacci 2014).
The geodatabase offers other benefits to housing large amounts of spatial data. Rather
than having many different files that need to be accessed for large projects, like wildlife tracking
or sensor records, the geodatabase contains all of the files required in one location which frees up
system resources. The geodatabase also provides several key elements that allow for better
functionality when dealing with spatial data. The creation of feature datasets and the
relationships between different feature classes within the datasets provides a perfect layout for
the data used in this project. Also, the ability to create domains that limit the inputs for a
11

particular field will help with data entry and error reduction. Finally, the geodatabase can allow
for quick updates based on the collected data using a data dictionary which aligns with GPS
receivers (Gonzalez-Tennant 2009).
The archaeological geodatabases give us a great basis for developing a presence/absence
geodatabase. Building a geodatabase for archaeological sites use temporal, spatial,
archaeological, and document data. For this presence/absence wildlife geodatabase there is also
temporal, spatial, and instead of archaeological the presence/absence data. The use of the Entity
Relationship Diagram (ERD) is vitally important to the creation of the geodatabase. Special
attention must also be given to geometry of the data because of its spatial nature and the type of
points, lines, or polygons input in a particular field. (Mocanu and Velicanu 2011; Breunig et al.
2016).
The design of the geodatabase must also take into account what the geodatabase will be
used for, as well as who will be using it. This project deals with researchers at the USGS which
is a federal agency. While not specifically designed for the entire agency some needs based
assessment can be used. Some federal agencies have strained to change from simple data
collection and storage to “a more collaborative system of data management” (Smith et al. 2015).
Many discussions where had with Dr. Beever on what the geodatabase design and usage would
encompass. The use of Smith et al. (2015) to better understand what questions to ask allowed for
a better understanding of how to construct the geodatabase for long term use.
2.2. Pika Habitat and Climate Change
North American temperature is expected to rise by more than 2°C by the middle of the
21
st
century and more than 4° by the end of the century, which will lead to changes in climate on
a micro and macro level (IPCC 2015). Some of the initial areas expected to be impacted by these
12

temperature changes are mountainous, or montane, regions due to the fact that temperature can
change rapidly with elevation. These montane habitats support a wide variety of species of plants
and animals, including the pika (Beniston 2003).
Species, including pika, that subsist exclusively in these montane areas have specially
adapted traits that allow for them to survive in these select environments (e.g. high body
temperature and low thermal conductivity). These traits have evolved previously by climatic
pressure forcing species to higher elevations and into these island habitats (Smith and Weston
1990; Beever et al. 2010). However, these selector traits were generally produced by long term
adaptation to climate factors rather than short rapid changes. Selector traits normally are formed
over the course of thousands of years and with the much more rapid change of climate
anticipated in the coming years, those selector traits will not change fast enough for some
animals to adapt. For pika, and other high montane species, this rapid and changing climate in
their habitable areas is unlikely to allow for the similar dispersal or adaptations seen in previous
shifts of habitat range (Barnosky and Kraatz 2007).
Talus are described as “piles of broken rock fringed by suitable vegetation” with rock
pieces ranging in size from 0.2m to 1m in diameter (Smith and Weston 1990). Along with
selecting talus locations, pika have adapted specific traits to live in these island habitats at high
elevation. A high natural body temperature (mean=40.1°C), with a limited range of
thermoregulation, and a low thermal conductance allow them to survive in these talus
(MacArthur and Wang 1973; Smith and Weston 1990). Although these traits have provided the
pika with the ability to survive the climate in high montane areas, they have put limitations on
pika’s ability to adapt to sudden changes in their climate. The natural high body temperature is
close to their high lethal limit and makes the pika susceptible to extreme temperatures, especially
13

higher than normal temperatures during summer months (Otto et al. 2015; Stewart et al. 2015;
Beever et al. 2003; 2010; Wilkening et al. 2011).
As a result of expected rapid change in temperatures at high elevations, the range and
suitable habitats for pika are likely to greatly diminish. There have been numerous studies of
climate related change on pika habitats using many different models (Mathewson et al. 2017;
Beever et al. 2003; Beever et al. 2010; Beever et al. 2011; Beever et al. 2013; Wilkening et al.
2011; Millar and Westfall 2010; Calkins et al. 2012; Calkins 2010). These all shows that with
expected temperature gains, suitable pika habitats will decrease in total area as temperatures rise.
Calkins (2010) showed that with a temperature increase of 7°C, expected habitable areas would
decrease by 95% in the Rocky Mountain lineages. Pika habitat loss is expected, and in some
cases already observed, to be especially prevalent in the lower elevation habitats, as well as
toward the southern boundary of their North American range (Beever et al. 2003; Beever et al.
2010).
While climate related extirpations are on the rise, other pika extirpations of historically
habitable sites were seen on the northern latitude of the Great Basin. These extirpations had
strong correlations with habitable talus located at lower elevations, in livestock grazing areas,
near roadways, and limited nearby habitable talus for dispersion (Beever et al. 2003). In addition,
surveys conducted in the 2000’s revealed extirpations exclusively located on the southern edge
of the pika Great Basin range. These extirpations were much more strongly correlated with rises
in overall temperature at those surveyed locations than with the previously seen extirpations at
the historical sites (Beever et al. 2010). So, while in recent years there has been strong evidence
to support temperature being the cause of extirpations, there are other factors to consider when
solely analyzing presence/absence at a given talus location.
14

Due to their restricted habitat (i.e. high mountain talus), as well as their relative ease of field
documentation, pika are a prime candidate for climate impact studies. As Beever et al. (2010)
note, the ability to document pika presence, by experienced observers, is as high as 95.9%. This
allows for a highly effective field study in which it is easy to detect a change in presence,
absence, or extirpation from a particular given site or talus. This project allowed for previously
collected pika presence, absence, and extirpation data to be housed and associated with specific
talus.
The data that has been collected in previous field seasons also includes temperature data
at several pika presence sites collected over multiple years. The inclusion of this data allows for
analysis of pika habitats and persistence, but also allows for a collection of temperature data to
be accumulated for montane/alpine conditions that may not be easily tracked otherwise. As
temperature is a vital determinate of pika presence, having this data available for multiple field
sites will dramatically increase the ability to predict extirpation (Beever et al. 2003; Beever et al.
2010; Beever et al. 2011; Beever et al. 2013).

15

Chapter 3 Methods
This chapter describes the methods used to complete the geodatabase project presented in this
paper. Section 3.1 is an overview of the design of the geodatabase and an overview of the
methods. Section 3.2 looks at the data, software, and hardware required for the completion of the
geodatabase, as well as future field work. Section 3.3 discusses the timeline for completion of
this project. Section 3.4 discusses the step by step process for completing the geodatabase and
data importation.
3.1 Methods Discussion
The design for this project is laid out in the following section. Section 3.1.1 discusses the
geodatabase design. Section 3.1.2., provides an overview of the data used in this project and
Section 3.1.3 discusses the detail of the relationships that were built in the database.
After completion of the database, data were formatted from previously housed Excel
spreadsheets and exported KML files. These data were then imported into the proper feature
classes within the database. Upon completion of the data import, digitization of numerous
PATCH polygons was completed to fill in some of the missing information for that feature class.
Finally, some queries were run to test for proper importation and housing of imported data.
3.1.1. Geodatabase Design
Initial geodatabase design was done via several personal communications with project
researchers to understand what data were already collected and what would be needed for future
research. While initial design sessions resulted in the ERD seen in Appendix A. This ERD was
developed through several discussions with Dr. Beever and before his field data were turned over
to the author (due to field season scheduling conflicts). As such, it was initially much broader
16

and the geodatabase design attempted to contain many more attributes in each Feature Class than
was finally deemed necessary. Once the data were procured and visual inspection of the data
were completed, some of these attributes changed within their respective feature classes to
accommodate formatting of the actual data. The Final ERD can be seen in Figure 2. Many of
these changes were made due to data which were not always consistent between field seasons.
Also, several of the feature classes, Survey, Mountain range, County, and State, were excluded
after further discussions and during examination of the data due to the fact that they could be
housed within an attribute of an alternate feature class and were not required to be separate
classes at this time.
The design of the initial structure is crucial to the success and usability of a database. The
building of a geodatabase requires knowing your audience and knowing your data (Smith et al.
2015). This project’s main audience are current USGS researchers. While audience is an
important factor so are the data. The data for this project involves years of presence/absence
points for pika at specific talus among the Great Basin region of North America. Included in the
data are presence/absence points and temperature/relative humidity from several of the field
study talus. Knowing that this initial project seeks to answer questions and queries about specific
talus and sites of extirpation makes the database design much easier.
This project requires a geodatabase that houses the spatial data from the digitized
polygons (i.e. the talus locations), the spatial points taken during site surveys for
presence/absence of pika, and temperature records taken using sensors buried at specific
locations in chosen talus. Among these data there are different types of attributes for each set of
surveys. Attributes of the digitized talus include: surveyed, possible habitable talus for survey,
17

Figure 2. Final ERD with changes to Feature Classes and Attributes
18

deemed unsuitable, persistent, extirpated, and functionally extirpated. Point data associated with
each polygon include: temperature recording points, which include year around temperature data,
and presence points with specific attributes the Dr. Beever’s research has shown to indicate high
probability of extant pika (i.e. ) hay piles, calls, etc.) (Beever et al. 2003). The development of
feature classes for geodatabases, allows for specific relationships to be built between the
different existing features (Gonzoles-Tennant 2009). The ERD in Figure 2 shows feature classes
for talus (polygon), temperature (point), presence (point) data, and their relationships.
3.1.2. Data Overview
The data used in this project was all provided by Dr. Beever. It includes point data for 40
individual sites. These sites are several historically known pika extant locations along with sites
that Dr. Beever has added to the field seasons through the years. They are provided as points and
then are buffered to 3000 meters, which allows Dr. Beever to know which patches need to be
field surveyed.
The next data set are patches. These areas encompass an area of talus that is continuously
connected. The patches are polygons and are linked to the site data based on an attribute and a
relationship class (which is further discussed in Section 3.1.3 and in Figure 4). The patches
house many attributes that are recorded during field season surveys which includes surveyed
status, pika population, mountain range, land manager, etc.
The final two datasets are point data which included iButton temperature/relative
humidity sensors and the presence/absence points recorded from the field surveys. These data
contain specific point locations of iButton sensors and pika presence/absence points within each
patch polygon. They are also nested in the geodatabase using a relationship class (Figure 4).
19

3.1.3. Detail of Relationships
Construction of the database occurred by creating a Feature Dataset with multiple Feature
classes within the dataset (see Figure 3). The feature classes included point feature classes
DETECTIONS (the presence/absence data), the iBUTTON (temperature/relative humidity data),
and SITES (overall site data). Also included was the polygon feature class PATCH (patch/talus

Figure 3. Feature Dataset, Feature Classes, and Joins

data). These were also given relationship classes between PATCH and DETECTIONS, SITES
and iBUTTON, and SITES and PATCH to allow for understanding the nested behavior of each
different type of feature class in relation to the others. Using Figure 4 as an example, Arc Dome
is a Site in the SITES feature class and ARDO1f, ARDO1c, and ARDO1h are talus patches
within the PATCH feature class. Within the ARDO1f patch there is a PRESENCE/ABSENCE
point 1447. These relationship classes were then used to create the corresponding joins between
the feature classes for querying purposes. Initially all database construction was completed
20

within ArcCatalog 10.5, with some subsequent editing done within ArcMap 10.5 and the Catalog
extension housed within.

Figure 4. Depiction of Relationship Classes in ArcMap
3.2. Data, Software, and Hardware Requirements
The data, software, and hardware requirements are listed in the following subsections
3.2.1, 3.2.2, and 3.2.3 respectively. 3.2.4 discusses timeframe for project completion.
3.2.1. Data Requirements
The data requirements for this project rely almost exclusively on data provided by Dr.
Erik Beever. These data are mostly point features that include presence/absence GPS points
taken over numerous field seasons. Other data that was included for this project are polygons and
alternative point data that have been housed on www.Caltopo.com. These data are a compilation
of previously digitized talus and point data representing both sites and presence/absence points
(see Figure 5).
21

3.2.2. Software Requirements
This project used several different pieces of software to complete. Initial data were
housed in Microsoft Excel and www.Caltopo.com. Microsoft Excel was also used to format the
field data into the appropriate fields for proper importation into the geodatabase. The polygon
and some point data were housed on www.Caltopo.com before exportation into the geodatabase
via KML files. The geodatabase was built in ArcCatalog 10.5 and ArcMap 10.5 was used to run
queries and verify all data were imported and house correctly.
3.2.3. Hardware Requirements
The software for this project was all used on Windows PCs and several different
handheld GPS units were also used during the field collecting process.
Figure 5. Example of digitized talus on www.caltopo.com

22

3.3. Timeline
The timeline for the completion of this project was structured roughly as seen in Table 1
below. Overall from start to finish the project took roughly 7 weeks to complete starting at data
collection and creation of the database structure. This was an easier than expected task, due to all
data relevant to the project being accessible. However, there was some work needed to develop
the database structure and multiple conversations were had to complete this process.
The formatting/standardization of the field data were a process that was much more
involved than originally planned. Due to the fact that data collection has been happening over the
span of roughly 20 years, much of the data were housed in different formats. This caused a
significant amount of formatting to allow for proper importation into the database. The
formatting of these data pushed back the original timetable considerably.
The formatting of data were followed by data exportation from www.Caltopo.com and
importation into the geodatabase. There were differences between the importation methods used
for the data from Excel versus the data from www.Caltopo.com in KML format and is discussed
in Section 3.4.
Finally, many remaining talus patches were digitized to allow for better proof of concept
of the geodatabase. Much of the PATCH data were not completed and was required to be added
at this point to fully allow for querying of more than one Site. By no means were all of the
remaining talus areas digitized at this point and further time was required to complete this
section.

23

Table 1. Project Timeline
Week 1-2 Weeks 3-6 Weeks 6-7 Week 8
Task 1
Creating database
structure

Task 2
Formatting
field data

Task 3
Exporting/importi
ng datasets

Task 4 Digitizing polygons
3.4. Detailed Methods
This section discusses in detail the different processes taken after design and build of the
geodatabase. Section 3.4.1 discusses the process used for data standardization and importation
into the geodatabase. Section 3.4.2 discusses the digitization of additional talus polygons and
3.4.3 discusses the queries used to test the geodatabase once completed.
3.4.1. Data Standardization and Importing
Field detection data for Dr. Beever’s Sites have been collected on and off since the mid
1990’s and were all previously housed in Excel spreadsheets with some years combined and
others housed separately. These data were collected as point data and attributes were assigned
based on field collection (i.e. hay piles, call, sighing, etc.). One issue with these data were that
they have been collected and formatted previously by different people and as such not all data
were collected or housed with the same attributes. The inconsistency with these data coupled
with the upwards of 2000 records made for many manual corrections, as well as many null
values for records in which the data were not captured. One example of some Detection data is
seen in Figure 6. The spreadsheet housed attributes for each point including Site, Date, Time,
Pika detected, etc.
24

Figure 6. Unformatted Detection Data (Latitude and Longitude Redacted)
25

Beginning with one spreadsheet that represented the most complete record of attributes,
the initial formatting was constructed. The easiest and most consistent process was found to
format/copy all records into one single CSV file for importation rather than attempting to import
multiple CSV files into the feature class. This was due to formulas and other formatting within
the original Excel spreadsheets. Each column was titled appropriately based on the naming
convention in the DETECTIONS feature class and formatted for the correct type of data (i.e.
Double, Integer, Text, etc.). These names were all based on Dr. Beever’s collection records and
can be seen below in Figure 7. Once all records were copied into the proper columns the
spreadsheet was saved as a CSV.

Figure 7. Detection Feature Class Formatted Fields
26

In one instance of the previously saved detection data, the XY coordinates were saved in
a separate location. To create the necessary CSV file, the data from the two separate spreadsheets
was merged into one workbook. Once the data were within the same workbook a VLOOKUP
was used to find the matching data based on the location name. The matching data were copied
to the proper location within a single spreadsheet and then formatted correctly to allow for
proper importation into the database.
Importing the CSV files into the geodatabase was done in a couple different steps. In
ArcMap, the Add Data tool was used to add the CSV files to a blank map. The CSV files initially
come in as tables with no geographic data displayed. Right clicking and selecting view XY data
allows for selection of the columns associated with Longitude and Latitude in decimal degrees,
as well as elevation data, and the proper datum (in this case WGS84). Once the XY data is
displayed the whole table can be exported as a feature class. This feature class can then be
imported into the premade feature dataset already housed in the geodatabase, with each column
creating an attribute corresponding to the attribute desired in the geodatabase feature class. This
technique was used for importing the Presence/Absence, Site, and iBUTTON data.
Data for the talus polygons was housed two separate ways. The polygons of the talus
themselves were housed via www.CALTOPO.com. This website has high resolution imagery
that allows for digitization of polygons and uploading points to have the visual data associated
with the detections and surveyed talus. This data is available to export via KML files which was
done for all previously digitized talus. The polygons were then imported into ArcMap via the
KML to Layer tool, which creates a feature class from a KML file. Upon completion of this step
this feature class was imported into the PATCH feature class in the feature data set only
importing the polygon data and the color data, as this was how the surveyed/unsurveyed
27

Figure 8. Example of Attributes for PATCH Feature Class joined with
SITES Feature Class
28

polygons were distinguished. Once imported, the CSV file housing the remaining data were
imported with the proper attributes for each talus polygon (see Figure 8). These attributes
coupled with the Joins between the different feature classes allow for substantial querying of
collected field data.
A table for the collected data by each iButton sensor was added in the final step. This
data is a simple Excel spreadsheet containing temperature and relative humidity that is collected
every four hours from placement of the sensor until retrieval or end of life. This data were
included in the geodatabase in a table simply to show proof of concept that the data could be
housed if necessary and for querying purposes for this project. In the future it may be more
beneficial to house this data externally due to the sheer number of records that each sensor
generates (upwards of 2000 records per sensor per year). The table was then joined with the
iButton point feature
3.4.2. Digitizing Polygons
The final step to completing the geodatabase was the continuation of digitizing the talus
that had been surveyed or may be surveyed in the future. This process was relatively simple once
the feature class for PATCH was complete. Using the create feature tool, an Esri Basemap with
satellite imagery, additional talus patches were digitized. This process included choosing what
type of patch was to be digitized, i.e. Surveyed, Unsurveyed – Potential, and Unsurveyed – Iffy
(this designation was included from previously digitized data for consistency). Patches were
selected in discussions with Dr. Beever and examining previously digitized areas for similar
categorized areas. The selected patch was then digitized by simply encompassing the patch area
in the polygon on top of the imagery (see Figure 9). The patch data does come with a caveat, the
patch data included for this project in no way encompasses all patch data that is needed for
29

complete querying of presence/absence points. It was required to digitize more patch areas
because of how few had been previously completed. However, with the already completed talus
patches, along with the patches digitized for this project, there is ample data for proof of
concept/design.
One difficulty that was uncovered in digitization and importation of the talus was some
attributes that had to be manually altered to allow for proper querying. Due to www.Caltopo.com
only being able to record polygons and a color (i.e. Surveyed, Unsurveyed, etc.) and not being
able to house attributes such as the site relationship for the PATCH polygons or the patch
relationship to the PRESENCE point data, much of this was required to be added manually in the
Figure 9. Example of Digitization of patches in ArcMap
30

database. This required inspecting each site area and adding the proper Site number to the proper
attribute field in the PATCH feature. The same was required for each Presence/Absence and
iButton point feature that fell within a digitized PATCH. This required some effort because some
sites have many different patches and each needed to be done individually. Because of this
limitation it was also decided to digitize additional talus patches in ArcMap rather than on
www.Caltopo.com and export them into the geodatabase.
3.4.3. Queries
Finally, once all data were imported, data verification occurred by testing several queries
(see Table 2). These queries were built based on conversations with Dr. Beever regarding his
future research questions. The results of these queries are discussed in full in the following
chapter.
Table 2. Queries for Testing Geodatabase Functionality
Queries Questions Asked Question Drivers
Query 1 What records were compiled during the
2016 field season?
Will allow for better planning of
future field seasons.
Query 2 What are the minimum and maximum
elevations of detected pika?
Elevation is an important factor for
ongoing pika extirpation research.
Query 3 How many potential patches need to be
surveyed?
Will focus field season surveys
towards patches in need of survey.
Query 4 What iButton locations have pika
occupancy from 2010 and 2017?
Example of basic analysis of
extant populations.
Query 5 What is the highest temperature recorded
by an iButton Sensor?
Allows for understanding of where
highest temperature was recorded.
Query 6 What sites had a decrease in pika
population from the 1990’s to the
2000’s?
Example of basic analysis of
population decreases between
surveyed patches.

31

Chapter 4 Results
This chapter discusses the results seen in the completed database. Section 4.1 describes all final
data and relationships used, as well as numbers of records included in the final geodatabase
build. Section 4.2 provides the answers to the results of the queries. Finally, section 4.3
summarizes all results for this project.
4.1. Final Design and Data
Section 4.1.1 looks at the final geodatabase design. Section 4.1.2 looks at the final data
and records housed in the geodatabase after importation and digitization.
4.1.1. Final Design/Schema
The final design took the initial ERD and changed certain feature classes and attributes to
account for collected data and ease of querying. The final ERD for this project is shown in
Figure 3 above and includes the feature classes and relationships deemed necessary for database
design. As discussed in the previous chapter the end design for the project required several
changes to allow for proper importation of the data. The original schema was designed before
taking possession of any data and was based on initial understanding of database needs. Once the
data were in hand, the understanding of how to properly house it in the geodatabase was better
realized, leading to the changes previously seen.
4.1.2. Geodatabase Data and Records
At the completion of this project, the feature classes and tables did not house all of Dr.
Beever’s data. There are still many talus polygons to digitize, detection and temperature data to
import, and possible future survey sites to be identified and imported. However, for this project
40 records were imported for the SITES feature class, 292 records for the iBUTTON feature
32

class, 405 records that combine previously digitized and current digitization for the PATCH
feature class, 1602 records for the DETECTIONS feature class, and 4096 records for the
temperature and relative humidity data for one iBUTTON location (see Figure 10 and Table 3).
At this time, the SITES and iBUTTON feature classes house all locations and data known, while
the PATCH and DETECTIONS feature classes and the SENSORS table house some, but not all,
of the data collected.
Table 3. Final Count of Imported Records
Features Type # of Records
Sites Point 40
Detections Point 1602
iButton Point 292
Patch Polygon 405
Temperature/Relative Humidity Table 4096

33

Figure 10. Visualization of all data showing Pika habitat in the study area of the Great
Basin Region (Patch/Polygon data is not visible at this scale)
34

4.2. Query Results
Upon completion of all data importation, the previously listed queries (seen in Table 2)
were done using ArcMap and the Selector tool to confirm proper importation and containment.
The Tables below show the queries with their results both graphically and in table format. The
individual queries showed all results as expected and allowed for nesting queries as needed.
These were used to accommodate some of the queries that involved calculating the maximum or
minimum values. Once imported into the geodatabase the spatial nature of the data becomes
more apparent. The ability to spatial analyze these data along with the other attributes provides
an enhanced understanding of how all the pieces of the field data interconnect.
Query 1 returned all values in the Detections feature class that were recorded in 2016.
This was 120 results and they are visualized in Figure 11. Queries such as this will allow better
understanding of field collection seasons and will allow for a better understanding what sites and
surveys were recorded in a given year. Seeing these spatially also allows for a specific and
concise plan for future field seasons.
Query 2 shows how researchers could better understand what elevations pika are
currently inhabiting and where they have extirpated from previously known habitat. The ability
to look at how far pika have retreated upwards or extirpated from specific locations, based on
elevation, will be valuable in future research. These queries returned both older records from the
mid 1990s with a minimum value of 1258.8m and a maximum of 3557m (see Figure 12 and
Figure 13). Further analysis regarding where these detections were recorded, as well as if any
have been recorded more recently, would allow for better understanding of pika persistence
today.

35

Query Number Query Results
Query 1 What records were surveyed during field
season 2016?
120
Figure 11. Query and Results showing locations of presence/absence points taken during
the 2016 field season.
36

Query 2 What are the minimum and
maximum elevations of
detected pika?
Min: 1258.8m
Max: 3557m

Query of Minimum Elevation with
Detected Pika
Query of Maximum Elevation with
Detected Pika
Figure 12. Queries for finding the maximum and minimum elevation of pika
detections
37

In Query 3, the patches that still need to be surveyed in a given site were queried. This
will allow users to plan future field seasons more efficiently. By better understanding and
spatially visualizing when and what areas have been surveyed recently, new survey areas can be
more effectively planned for a given year. Currently the database showed 71 unsurveyed
potential patch locations that need to be surveyed (see Figure 14). As discussed previously, not
all talus areas have been digitized. Once this is completed this query will generate increased
results.
Results from Minimum Elevation Query
Results from Maximum Elevation Query
Figure 13. Results for finding the maximum and minimum
elevation of pika detections
38

Query 3 How many potential patches
need to be surveyed?
71

Figure 14. Query and Results of Unsurveyed Potential Patches
39

The data for the iButton locations is particularly interesting as it adds relative humidity
and temperature data for each talus patch where the sensors are placed. Also, all talus data (i.e.
presense/absence, etc.) is collected with each placing and collection of the sensor data, which
provides additional valuable information. Query 4 finds the results of pika persistence from
iButton sensor locations from 2010 survey to 2017 survey. The query returned 77 of the iButton
locations as having persistent pika (see Figure 15).

Continuing with iButton queries, but this time focusing on the sensor table, Query 5 asks
what the highest temperature recorded in the sensor table was. This resulted in a record for
Query 4 What iButton locations have
pika occupancy from 2010
and 2017?
77
Figure 15. Query and Results of iButton
Locations with Pika Occupancy in both 2010
and 2017
40

29.04°C (see Figure 16). A caveat with this query is that only one set of sensor data were
imported from only one sensor. This was due to the large number of records that are stored by
each sensor. As stated above, one sensor for a 2 year period generated over 4000 records. This
data is also not readily available to the author at this time and so one sensor was included to
show proof of concept.

Query 5 What is the highest
temperature recorded by an
iButton Sensor?
29.04°C

Figure 16. Query and Results of Highest Temperature
Recorded
41

The final query was run to search how many sites saw a decrease in pika population from
the 1990s levels to the 2000s levels. The results returned 11 sites that have seen pika population
decreases between the two time periods (see Figure 17 and 18). This final query was run as a
more personal exercise to see if pika populations were in fact reducing during the study
timeframe. While some of the sites did see reductions in pika populations, it should be noted that
others saw increases. Some of these variations could simply be due to factors surrounding the
survey, but the number of sites that saw an increase was far outweighed by the number that
decreased.

Query 6 What sites had a decrease in
pika population from 1990’s
to the 2000’s?
11

Figure 17. Query and Results of Sites that Lost Pika Population between the
1990's field seasons to the 2003-2008 field seasons
42

Figure 18. Results of Query 6
43

4.3. Summarizing of Results and Errors
As can be seen by the previous sections, the database was successfully populated to allow
for querying of the data. The importation of many of the Detection records and creating the
relationships with the attributes and the sites will greatly benefit future research. The ability for
users to visualize the relationships between the Detections, the Patch, and the Site will allow
them to greatly increase the speed of analysis and presentation of results.
The building of the relationship classes between Sites, Patches, Detections, and iButton
sensors will allow for future collaboration in this research and understand the nesting concept of
this work (see Figure 19). The hope is that future field seasons can be easily planned using this
geodatabase and the contained records. The visualization of all data from the macro (Site level)
to the micro (Presence/Absence points) allows the users to intuitively see the relationships.
Figure 19. Relationship of nested features example showing
the Arc Dome Site with the ARDO1e Talus Patch related to
it with the Presence/Absence Point 1442 related to both.
44

A final example, which can be seen below in Figure 20, shows a complete view of Site
11 with all associated feature classes. The relationships for this site are shown above in Figure
19, which represent Site 11, the Patches associated with Site 11, the Detection points for each
Patch, and the iButton sensors placed in each Patch. This was the goal at the beginning of this
project and so the belief is that this will ease the analysis of collected data and greatly increase
the research capabilities. With the goal of housing all data in one database met, the ability to
continue with further spatial analysis on all data will be a huge benefit to current and future
research.
There were several errors that were found when importing the data. These revolved
around the formatting of each attribute in Excel and how they were saved in the CSV file. One
example was several of the attributes were initially built in the database as a Short or a Long type
but had to be adjusted once the data were imported. This required several iterations of
importation and reformatting the database to accommodate how the data were originally
formatted during the field collection.
One feature in ArcCatalog that the author found particularly helpful was the ability to
build a geodatabase schema based on the formatting of the import data. Gathering all the data
into the CVS file and formatting each attribute accordingly allows for ArcCatalog to build a
feature class schema based on the formatting from the CVS file. This allows for quick database
construction and eliminates the struggles of possibly building your feature class attributes with
the wrong type code.

45

Figure 20. Example of Site with all Feature Classes
46

Chapter 5 Conclusion
In this chapter the conclusions drawn from this project are discussed. Section 5.1 discusses the
lessons learned while completing this project. Section 5.2 discusses the uses and current
completion state of housing all the data in the geodatabase. Section 5.3 discusses the future work
related to this project.
5.1. Project Lessons
This study aimed to build a geodatabase to house the research collected by Dr. Beever’s
research project on pika and climate change. This project was originally much more complex
than the final design. Due to many of the time commitments to formatting and importing the
data, less time was spent on including all surveyed data and more on formatting the base data.
The underlying structure that was completed during this project provides a framework that
allows for expansion should others continue with this work.
The original complexity of the project was mostly due to limited understanding on how
the data were surveyed. This then led to the initial design of the database which proved much
more complicated than necessary, as well as created some more formatting issues to allow for
proper importation. The design portion would have been much better served to have allowed
time to gather all datasets and review them at once, rather than slowly piece them together over
several weeks. This probably could have been avoided but, due to overlapping deadlines for this
project and field season requirements, it was necessary.
Another challenge of this project was understanding exactly how all pieces of research
were connected. After several discussions with Dr. Beever and reviewing of all data this became
clearer and allowed for much faster geodatabase design. The ability for laymen to quickly
visualize and understand how the data fit together will allow for a greater collaboration on
47

research in the future. The visualization of the geographically nested data will benefit the project,
in the ability to quickly find points associated with each Site or Patch with relative ease. The
hope is that future field seasons can be easily planned using this geodatabase and the contained
records.
Additionally, much more time than was previously thought was needed to format the data
for proper importation. This did involve some querying within Excel to allow for the proper
attributes to be imported to the proper feature classes. In the future, this may be able to be
avoided by field collecting the data in slightly different format, or translating field work to the
Excel spreadsheets in a more efficient way. Much time was spent simply testing data importation
because of errors in proper formatting of the columns in the CSV files. Also as previously
discussed, it was also found to be much more efficient to simply format the CSV file and then
use that to create the formatting for each attribute in the feature class by importing the schema
(e.g. Double, Text, Float, etc.) which alleviated much of the formatting issues encountered early
in the project.
5.2. Current State of Geodatabase
Due to the ongoing nature of the pika research project, this geodatabase will more than
likely need to be updated frequently. The included records and schema for this project were an
attempt to show that using a geodatabase will greatly increase querying ability to quickly
understand which sites have been surveyed, when they were surveyed, and what the results of
those surveys. This process will also be enhanced due to the visual results provided by the
geodatabase platform. Previously it was required to comb through, query separate spreadsheets,
or rely on institutional knowledge, to understand what was needed to be accomplished each field
season.
48

While this project does not include all records that have been collected. The basis has
been provided to show how the data may be housed that will allow for speedy importation in the
future. With the addition of a data dictionary, the data collection to importation process would
speed up immensely and allow for much faster results each field season. Having a data dictionary
would be a great asset if implemented in future field seasons. The construction of a data
dictionary will eliminate all the need to format data and allow for concise and accurate collection
of the data. However, the data dictionary construction was not attempted during this project due
to time constraints.
5.3. Future Work
The future work for this project could be extensive. The data collection is currently
scheduled to continue for the foreseeable future and because of this, the need for this database
should be ongoing for years to come. With continuation of the data collection comes the
opportunity to ease the data collection methods. It would greatly speed up collection and post-
processing of the field data if some type of data dictionary could be used to collect the data in the
field. This could be set up on iPad’s using Collector for ArcGIS or via a Trimble GPS unit and
GPS Pathfinder. These hardware and software options would provide a valuable source of
collecting that would all but eliminate any formatting needs once the data were collected.
The use of a data dictionary would allow for coded values and other such methods to
limit certain inputs that are being used in the database. This would make for a simple easy
download and import into the geodatabase at the end of each field day or field session. Once this
system is implemented, the geodatabase would be much more up to date and alleviate the
formatting issues that plagued this project.
49

Another aspect that should be expanded on is the inclusion of more temperature data.
This data encompasses over 200 locations with accurate temperature and relative humidity for an
extended area of the western United States. The inclusion of all of these records will allow for a
valuable understanding of how climate is being affected at higher elevations over the course of
many years. These data can also be compared to data from temperature models to determine how
accurate they are at predicting temperature in montane environments.
Collected alongside the temperature data were also relative humidity data. These data
will also allow for further analysis of climate at higher elevations and allow for a more complete
picture of change, should it be shown to have occurred. The inclusion of relative humidity data
can greatly increase the spatial analysis that can be performed and correlated to burn areas and
further Dr. Beever’s research on climate change and impacts. The links between relative
humidity and fire danger have been shown to be particularly correlated with mean relative
humidity as the best overall predictor (Holsten et al. 2013). Many important analyses could be
completed using these two different data sets.
The inclusion of temperature and relative humidity data will cause some issues because
of the number of records generated. The sensors collect roughly 2000 records a year per sensor
which can become unwieldy. As seen in other environmental studies, other software such as
Loggernet may be better equipped to handle data volumes that are seen for these sensors (Gries
et al. 2016).
There is also an opportunity to house more data that has been collected at some, but not
all, of the Sites (i.e. survey specific data). This data includes, but is not limited to, wind speeds,
percent of area recently burned, percent of area with snow, animal sightings, and types of trees
and shrubs. While these attributes were not included for this project, it would be beneficial in the
50

long term to add this data, which would provide a more complete understanding of the surveyed
locations. The inclusion of landscape and other variables are often the best predictors when
trying to understand wildlife habitat (Carroll et al. 1999).
Finally the complete housing of all of research in one location will allow for a much more
thorough analysis of these data. With the ability to see all work in one location, and with a
streamlined data collection process, research will be able to continue in a much more efficient
way. It will also allow for the data to be presented and analyzed geographically. This will help to
understand and visualize where specific changes have occurred and provide a palette from which
to work from for future research.

51

References
Beever, Erik A., Peter F. Brussard, and Joel Berger. 2003. “Patterns of Apparent Extirpation
among Isolated Populations of Pikas (Ochotona princeps) in the Great Basin. Journal of
Mammalogy. 84, No. 1: 37-54. http://www.jstor.org/stable/1383633.

Beever, Erik A. Chris Ray, Philip w. Mote, and Jennifer Wilkening. 2010. “Testing alternative
models of climate-mediate extirpations.” Ecological Applications. 20, No. 1:164-178.
http://www.jstor.org/stable/27797796.

Beever, Erik A., Chris Ray, Jenifer L. Wilkening, Peter F. Brussard, and Philip W. Mote. 2011.
“Contemporary climate change alters the pace and drivers of extinction.” Global Change
Biology. 17: 2054-2070. Doi: 10.1111/j/1365-2486.2010.02389.x.

Beever, Erik A., S.Z. Dobrowski, J. Long, A.R. Mynsberge, and N.B. Piekielek. 2013.
“Understanding relationships among abundance, extirpation, and climate at ecoregional
scales.” Ecology. 94, No. 7: 1563-1571. http://www.jstor.org/stable/23596945.

Bruenig, Martin, Paul V. Kuper, Edgar Butwilowski, Andreas Thomsen, Markus Jahn, Andre
Dittrich, Mulhim Al-Doori, Darya Golovko, and Mathia Menninghaus. 2016. “The story
of DG4GeO – A service-bassed geo-database architecture to support multi-dimensional
data analysis and visualization.” ISPRS Journal of Photogrammetry and Remote Sensing.
117:187-205. Doi: 10.1016/j.isprsjprs.2015.12.006

Calkins, Michael T., Erik A. Beever, Kenneth G. Boykin, Jennifer K. Frey, and Mark C.
Andersen. 2012. “Not-so-splendid isolation: modeling climate-mediated range collapse of
a montane mammal Ochotona princeps across numerous ecoregions.” Ecography. 35 No.
9: 780-791. Doi: 10/1111/j.1600-0587.2011.07227.

Calkins, Michael T. 2010. “Modeling Climate Change Impacts on the American Pika Over the
Western United States.” Master’s Thesis. New Mexico State University.

Carroll, Carlos, William J. Zielinski, and Reed F. Reed F. 1999. “Using Presence-Absence Data
to Build and Test Spatial Habitat Models for the Fisher in the Klamath Region, U.S.A.”
Conservation Biology. Vol. 13, No. 6:1344-1359. http://www.jstor.org/stable/2641958.

Gonzalez-Tennant, Edward. 2009. “Using Geodatabases to Generate ‘Living Documents’ for
Archaeology: A Case Study from Otago Goldfields, New Zeland.” In “Current Research
in Australia and New Zeland.,” Historical Archaeology 43, No. 3: 20-37. Accessed July
11, 2016. http://www.jstor.org/stable/25617568.

Gries, Corinna, Mark R. Gahler, Paul C. Hanson, Timpoth K. Kratz, and Emily H. Stanley. 2016.
“Information management at the North Temperate Lakes Long-term Ecological Research
site – Successful support of research in a large, diverse, and long running project.”
Ecological Informatics 36: 201-208. http://dx.doi.org/10.1016/j.ecoinf.2016.08.007
52

Hafner, David J. 1993. “North American Pika (Ochotona princeps) as a Late Quaternary
Biogeographic Indicator Species.” Quaternary Research 39: 373-380.
DOI:10.1006/qres.1993.1044

Holsen, Anne, Anto Raja Dominic, Luis Costa, and Jurgen P. Kropp. 2013. “Evaluation of the
performance of meteorological forest fire indices for German federal states.” Forest
Ecology and Management. 287:123-131. Doi: 10.1016/j.foreco.2012.08.035.

MacArther, R.A. and L.C.H. Wang. (1974). “Physiology of thermoregulation in pika, Ochotona
princeps.” Canadian Journal of Zoology 51:11-16. https://doi-
org.libproxy1.usc.edu/10.1139/z73-002.

Mathewson, Paul D., Lucas Moyer-Horner, Erik A. Beever, Natalie J. Briscoe, Michael Kearney,
Jeremiah M. Yahn, and Warren P. Porter. 2017. “Mechanistic variables can enhance
predictive models of endotherm distributions: the American pika under current, past, and
future climates.” Global Change Biology. 23:1048-1064. DOI:10/1111/gcb.13454.

Millar, Constance I., and Robert D. Westfall. 2010. “Distribution and Climatic Relationship of
the American Pika (Ochotona princeps) in the Sierra Nevada and Western Great Basin,
U.S.A.; Periglacial Landforms as Refugia in Warming Climates.” Arctic, Antarctic, and
Alpine Research 42, No. 1: 76-88. http://www.jstor.org/stable/25642094.

Mocanu, Aura-Mihaela and Manole Velicanu. 2011. “Building a Spatial Database for Romanian
Archaeological Sites.” Database Systems Journal 2 No. 1 (2011): 3-12.
http://www.dbjournal.ro/archive/3/3.pdf

Moyer-Horner, Lucas, Paul D. Mathewson, Gavin M. Jones, Michael R. Kearney and Warren P.
Porter. 2015. “Modeling behavioral thermoregulation in a climate change sentinel.”
Ecology and Evolution 5. No. 24:5810-5822. DOI:10.1002/ece2.1848.

Otto, Hans W., James A. Wilson, and Erik A. Beever. 2015. “Facing a Changing World:
Thermal Physiology of American Pikas.” Western North American Naturalist. 75, No.
4:429-445. http://libproxy.usc.edu/login?url=https://search-proquest-
com.libproxy1.usc.edu/docview/1770937946?accountid=14749.

Rodhouse, Thomas J. Erik A. Beever, Lisa K. Garrett, Kathryn M. Irvine, Mackenzie R.
Jeffress, Michael Munts, and Chris Ray. 2010. “Distribution of American pikas in a low-
elevation lava landscape: conservation implications from the range periphery. Journal of
Mammalogy. 97, No. 5: 1287-1299. American Society of Mammalogists.
http://www.jstor.org/stable/40925736.

Romero-Lankao, P., J.B. Smith, D.J. Davidson, N.S. Diffenbaugh, P.L. Kinney, P. Kirshen, P.
Kovacs, and L. Villers Ruiz. 2014: North America. In: Climate Change 2014: Impacts,
Adaptation, and Vulnerability. Part B: Regional Aspects. Contribution of Working Group
II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change
53

[Barros, V.R., C.B. Field, D.J. Dokken, M.D. Mastrandrea, K.J. Mach, T.E. Bilir, M.
Chatterjee, K.L. Ebi, Y.O. Estrada, R.C. Genova, B. Girma, E.S. Kissel, A.N. Levy, S.
MacCracken, P.R. Mastrandrea, and L.L. White (eds.)]. Cambridge University Press,
Cambridge, United Kingdom and New York, NY, USA, pp. 1439-1498.
http://www.ipcc.ch/pdf/assessment-report/ar5/wg2/WGIIAR5-Chap26_FINAL.pdf

Rowe, Rebecca J. and Rebecca C. Terry. 2014. “Small mammal responses to environmental
change: integrating past and present dynamics.” Journal of Mammalogy. 95, No. 6: 1157-
11174. American Society of Mammalogists. DOI: 10.1644/13-MAMM-S-079.

Smith, Jordan W., William S. Slocumb, Charlynne Smith, and Jason Matney. 2015. “A Needs-
Assessment Process for Designing Geospatial Data Management Systems within Federal
Agencies.” Journal of Map & Geography Libraries 11, No. 2: 226-244.
DOI:10.1080/15420353.2015.1048035.

Stewart, Joseph A. E., John D. Perrine, Lyle B. Nichols, James H. Thorne, Constance I. Millar,
Kenneth E. Goehring, Cody P. Massing, and David H. Wright. 2015. “Revisiting the past
to foretell the future: summer temperature and habitat area predict pika extirpations in
California.” Journal of Biogeography 42:880-890.
http://wileyonlinelibrary.com/journal/jbi DOI:10.1111.12466.

Urbano, Ferdinando and Francesca Cagnacci, (eds.) Spatial Database for GPS Wildlife Tracking
Data. Springer International Publishing Switzerland, 2014. DOI: 10.1007/978-3-319-
03743-1_1

Wilkening, Jennifer L., Chris Ray, Erik A. Beever, and Peter F. Brussard. 2011. “Modeling
contemporary range retraction in Great Basin pikas(Ochotona princeps) using data on
microclimate and microhabitat.” Quaternary International. 235: 77-88. Elsevier. DOI:
10.116/j.quaint.2010.05.004.

54

Appendices
Appendix A – Initial Entity Relationship Diagram

Abstract (if available)

Abstract A United States Geological Survey (USGS) researcher has been studying impacts of climate change on American Pika (Ochotona princeps) from the mid 1990’s through 2017. This project aims to contribute to research on the American Pika by building a geodatabase to store and provide access to data on pika populations throughout the Great Basin region of the Western United States. The geodatabase contains pika presence and absence data for locations of talus, which includes habitat areas that have been previously surveyed or may be potentially surveyed in the future. The project used formatted data provided from field surveyed talus that have been digitized on www.caltopo.com, digitized new talus that have been more recently surveyed, and imported GPS points for presence/absence captured in Excel spreadsheets. ❧ The end result of this project was a geodatabase that housed presence/absence points, talus polygons, site locations, temperature sensor locations, and temperature/relative humidity data. Several queries were completed that show proper importation and relationships of all data. Working closely with project researchers, this study allows for database expansion as needed for future research needs. ❧ Studying presence/absence of American Pika allows for further understanding of climatic impacts in niche habitats that are especially susceptible to environmental change. This project also provides the opportunity for improved analysis and long term data storage relating to these presence/absence locations throughout the Great Basin region. The end result supports expansion of the database structure for future field seasons and data inclusion.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Creating a water quality geodatabase for the West Hawai‘i Island region

PDF

A comparison of GLM, GAM, and GWR modeling of fish distribution and abundance in Lake Ontario

PDF

A unified geodatabase design for sinkhole inventories in the United States

PDF

Preparing for immigration reform: a spatial analysis of unauthorized immigrants

PDF

Congestion effects on arterials as a result of incidents on nearby freeway: When should you get off the highway?

PDF

Creating a Web GIS to support field operations and enhance data collection for the Animal and Plant Health Inspection Service (APHIS)

PDF

GIS data curation and Web map application for La Brea Tar Pits fossil occurrences in Los Angeles, California

PDF

Building better species distribution models with machine learning: assessing the role of covariate scale and tuning in Maxent models

PDF

Building a spatial database of biochar research and practice with Web-GIS

PDF

Spatial and temporal patterns of long-term temperature change in Southern California from 1935 to 2014

PDF

A comparison of two earthquake events in the City of Downey: the Puente Hills and Whittier faults at 7.0 and 6.8 magnitudes

PDF

Analysis of future land use conflict with volcanic hazard zones: Mount Rainier, Washington

PDF

Preparing for the next major southern California earthquake: utilizing HAZUS with soils maps and ShakeMaps to predict regional bridge damage and closures

PDF

A Web GIS application for airport pavement management

PDF

Increase in surface temperature and deep layer nitrate in the California Current: a spatiotemporal analysis of four-dimensional hydrographic data

PDF

Scenario-based site suitability analysis and framework for biodiversity conservation: agricultural zone, Galapagos Archipelago, Ecuador

PDF

Exploring global natural disaster and climate migration data: a Web GIS application

PDF

Preparing for earthquakes in Dallas-Fort Worth: applying HAZUS and network analysis to assess shelter accessibility

PDF

Evaluating predator prey dynamics and site utilization patterns of golden eagles using resource selection modeling and spatiotemporal pattern mining

PDF

Climate conservation priorities: using MCDA to identify future refuge for the Joshua Tree Forest in California

Asset Metadata

Creator Burke, Kyle Krueger (author)

Core Title Building a geodatabase design for American Pika presence and absence data

School College of Letters, Arts and Sciences

Degree Master of Science

Degree Program Geographic Information Science and Technology

Publication Date 04/12/2018

Defense Date 03/21/2018

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

Tag American Pika,ArcGIS,climate change,geodatabase,GIS,Great Basin,niche habitat,OAI-PMH Harvest,Ochotona princeps,talus

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Ruddell, Darren Martin (committee chair), Kemp, Karen K. (committee member), Loyola, Laura Cyra (committee member)

Creator Email kyleburk@usc.edu,kylek103@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-7316

Unique identifier UC11672128

Identifier etd-BurkeKyleK-6241.pdf (filename),usctheses-c89-7316 (legacy record id)

Legacy Identifier etd-BurkeKyleK-6241.pdf

Dmrecord 7316

Document Type Thesis

Rights Burke, Kyle Krueger

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