University of Southern California Dissertations and Theses

Quantifying changes in monarch butterfly habitat in California

(USC Thesis Other)

Quantifying changes in monarch butterfly habitat in California

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

Quantifying Changes in Monarch Butterfly Habitat in California

by

Indu Hulisandra

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 Indu Hulisandra

To my family

iv
Table of Contents
List of Figures ............................................................................................................................. vii
List of Tables ................................................................................................................................ ix
List of Charts ................................................................................................................................. x
Acknowledgements ...................................................................................................................... xi
List of Abbreviations ................................................................................................................... xii
Abstract ...................................................................................................................................... xiv
Chapter 1 Introduction ................................................................................................................... 1
1.1 North American Monarch Butterfly.................................................................................... 2
1.2 Western Monarch Butterfly ................................................................................................ 4
1.3 Research Aim ...................................................................................................................... 6
1.4 Study Area .......................................................................................................................... 7
1.5 Previous Studies and Gaps .................................................................................................. 9
Chapter 2 Related Work .............................................................................................................. 11
2.1 Land Use Land Cover Changes ........................................................................................ 13
2.1.1. Overwintering Sites ................................................................................................ 13
2.1.2. Spring and Summer Migration Habitat ................................................................... 17
2.2 Weather Changes .............................................................................................................. 19
Chapter 3 Data And Methods ...................................................................................................... 24

v
3.1 Data and Methods Overview ............................................................................................ 24
3.2 Research Design................................................................................................................ 25
3.3 Data Description and Methods.......................................................................................... 26
3.3.1. Land Use Land Cover Changes .............................................................................. 26
3.3.2. Digital Elevation Model Map ................................................................................. 36
3.3.3. Weather Data .......................................................................................................... 38
3.3.4. Pesticide Use Report ............................................................................................... 39
3.3.5. Overwintering Sites and Yearly Count ................................................................... 44
3.4 Output Results and Maps .................................................................................................. 44
Chapter 4 Results ......................................................................................................................... 45
4.1 Land Use Land Cover Change Analyses .......................................................................... 45
4.1.1. NDVI Image Generation and Results ..................................................................... 46
4.1.2. NDVI Image Differencing and Results................................................................... 48
4.2 Pesticide Use Report Analysis .......................................................................................... 58
4.3 Weather Change Analysis ................................................................................................. 65
4.4 Conclusions ....................................................................................................................... 68
Chapter 5 Discussion ................................................................................................................... 71
5.1 Conclusion ........................................................................................................................ 71
5.2 Future Scope of Work ....................................................................................................... 73

vi
References ................................................................................................................................... 77
Appendix A: Monarch Sites and Boundary ................................................................................ 85

vii
List of Figures
Figure 1: Migration patterns and over wintering sites. Image source: The Xerces Society ............ 1
Figure 2: Monarch Breeding and Milkweed Records of the Western Region ................................ 5
Figure 3: Study area Monterey County at township level and Monarch winter sites ..................... 7
Figure 4: Monarch Yearly Count. Source: Thanksgiving Count Xerces Society .......................... 12
Figure 5: Landsat ARD Tile and Grid format. Source: USGS Landsat ARD .............................. 29
Figure 6: Landsat 4-7 Surface Reflectance Specifications. Source: USGS Landsat. ................... 30
Figure 7: Landsat 8 Surface Reflectance Specifications. Source: USGS Landsat ....................... 30
Figure 8: Composite band image of Landsat ARD 1994 .............................................................. 32
Figure 9: Out of Range NDVI values ........................................................................................... 34
Figure 10: NDVI image and NDVI differencing process ............................................................. 36
Figure 11: Sample of daily weather data of the Monterey county from 1997-2016. .................... 38
Figure 12: Glyphosate application from 1991 - 2015 ................................................................... 42
Figure 13: PUR Table in ArcMap ................................................................................................. 43
Figure 14: NDVI Image Reclassified 1994 and 1998 ................................................................... 48
Figure 15:NDVI Image Reclassified 2002 and 2006 .................................................................... 49
Figure 16: NDVI Image Reclassified 2010 and 2014 ................................................................... 50
Figure 17: NDVI Image Reclassified 2018 .................................................................................. 51
Figure 18: NDVI Images of 2010, 2014 and NDVI difference image of 2010-2014 ................... 53

viii
Figure 19: NDVI Difference Image .............................................................................................. 55
Figure 20: NDVI Difference Image .............................................................................................. 56
Figure 21: NDVI Difference Image .............................................................................................. 57
Figure 22: Glyphosate application in sections and towns based on crop type.............................. 64
Figure 23: Glyphosate application in sections and towns based on crop type.............................. 65

ix

List of Tables
Table 1: Data, Type and Source .................................................................................................... 25
Table 2:NDVI Pixel Range and Categories .................................................................................. 33
Table 3: NDVI Class and Land Cover Percentage ....................................................................... 47
Table 4: Crop Type and Pounds of Chemical Used ...................................................................... 62

x
List of Charts
Chart 1: Land use land cover percentage ...................................................................................... 47
Chart 2: Land use land cover change in percentage ..................................................................... 58
Chart 3: Crops and year wise pesticide glyphosate applied .......................................................... 61
Chart 4: Pounds of Glyphosate applied each year ........................................................................ 62
Chart 5: Highest temperature recorded monthly since 1997 to 2016 ........................................... 67
Chart 6: Monarch Yearly Count. Source: Xerces Society............................................................. 68

xi
Acknowledgements
I am grateful to my mentor Dr. Jennifer Bernstein, for the advice, suggestion, patience,
and motivating me time and again. I would like to thank all the faculty members at Spatial
Sciences Institute who have taught me during my course work, and my committee members Dr.
Darren Ruddell and Dr. Andrew Marx for providing the inputs for this thesis. I am grateful for
the monarch butterfly data provided to me by Xerces Society.

xii

List of Abbreviations
ARD Analysis Ready Data
CDL Cropland Data Layer
COMTRS County Meridian Township Range Section
DEM Digital Elevation Model
DPR Department of Pesticide Regulation (State of California)
GDD Growing Degree Days
GIS Geographic Information Science
GM Genetically Modified
GLM Generalized Linear Models
HSM Habitat Suitability Modeling
ID Identification
MBBR Monarch Butterfly Biosphere Reserve
MLMP Monarch Larva Monitoring Project
MRLCC Multi-Resolution Land Characteristics Consortium
NCDC National Climatic Data Center
NDVI Normalized Difference Vegetation Index
NLCD National Land Cover Database

xiii
PCA Principal Component Analysis
PLSS Public Land Survey System
PUR Pesticide Use Report
UC IPM University of California Integrated Pest Management
USDA United States Department of Agriculture
USGS United States Geological Society
WMTC Western Monarch Thanksgiving Count

xiv
Abstract
Western Monarch butterflies in North America have seen a steep population decline from
1990 to 2019 due to climate change and land cover changes in their overwintering sites as well
as the decline of milkweed plants in the breeding grounds due to pesticide glyphosate
application. Since 1990, North American monarch populations have declined by 68 percent
overall, and western monarchs have seen nearly a 90 percent decline, reducing the population of
monarch butterfly from more than one billion count of butterflies to less than two hundred
million population size (CBD 2018). This project quantified land use land cover changes from
1990-2015 in Monterey county to identify possible impacts to Monarch butterfly habitat. The
land cover and land use change were analyzed using Landsat ARD imagery from 1994, 1998,
2002, 2006, 2014 and 2018, using the NDVI (Normalized Difference Vegetation Index)
differencing method. The NDVI images for each year were classified into 5 categories based on
the spectral index. The results indicate changes in the land use land cover, however, there was
not significant change in any class. Additionally, this project analyzed the application of the
herbicide glyphosate and weather changes for 1990 – 2015. The weather data analysis did not
show any inclement weather which might be unsuitable for monarch winter survival and the
pesticide use report showed that grapes and uncultivated agricultural land used high amount of
the herbicide glyphosate each year. The application increased between 1996 and 2000, and again
it is increasing from 2004 as some of the crop types are using more herbicide. The crucial
outcome of this study indicated that the monarch population count increased when there was
reduction in pesticide use the previous year and vise-versa, and increase in moderate and healthy
vegetation. This might indicate the milkweed might have been abundant during the spring and

xv
summer migration for the monarchs to thrive. Thus, study indicate use of pesticide glyphosate
affects the monarch butterfly population, whereas, the other factors such as land cover change,
and weather is not affecting western monarch population.

1
Chapter 1 Introduction
The North American monarch butterflies are a migratory species which travel between
winter sites in Coastal California and forests in Central Mexico every winter. These winter
habitats have optimum conditions for the monarchs to survive the cold months, which are
warmer and moderate, unlike the weather conditions in the rest of continental US. Monarchs
prefer warmer weather conditions as they remain in a diapause state (where their reproduction
cycle is in a suspended state) till the end of winter and the start of spring season. Monarchs begin
their migration in coastal California and the forests of Central Mexico and travel north to the
southern Canadian border during spring and summer (Fig 1). During their spring and summer
migration, they come out of the diapause state and start their reproductive cycle. During this
Figure 1: Migration patterns and over wintering sites. Image source: The Xerces Society

2
time, they use milkweed plants for food and to lay their eggs. For all reproduction stages (egg,
larva, pupa, and adult) milkweed is the host plant. Since 1990, the monarch population has
declined by more than 80 percent. Habitat degradation and loss in wintering sites, unusual
weather changes, and the decline of milkweed plants due to use of herbicide glyphosate in
spring-summer habitats have led to the species’ decline.
This project aimed to quantify land cover changes in the winter habitat in the Central
coast of California from 1990 to 2018, specifically Monterey County. This study site was
selected because, the county coastline has significant number of over wintering sites, in open
spaces, close to the urban areas, and sites with proximity to agricultural areas. Monterey County
alone has 15 active monarch butterflies overwintering approximately (as not all 15 sites see
overwintering butterflies each year). The land use land cover changes in and around the winter
sites, changes in the weather patterns, and the application of the herbicide glyphosate use from
1990 to 2018 were analyzed using Landsat imagery (ARD 1994, 1998, 2002, 2006, 2010, 2014
and 2018) NDVI image differencing method, weather data, and PUR reports.
1.1 North American Monarch Butterfly
Monarch butterflies (Danaus plexippus) migrate to overwintering sites in Central Mexico
and coastal California and roost in large clusters on host trees during winter (Griffiths and
Villablanca 2015). There are two distinct populations separated by geography. Monarchs west of
the Rocky Mountains migrate to the California coast during their fall migration and spend winter
months in Monterey pine and Eucalyptus groves. Monarchs east of the Rocky Mountains travel
during fall to spend the winter months in the Oyamel fir forests of Central Mexico. This winter
sites were discovered by Dr. Fred Urquhart (Urquhart and Urquhart 1976).

3
The fall migration starts from late September and they reach and roost in the winter sites
in coastal California and Central Mexico by November. During fall they enter reproductive
diapause state (Anderson and Brower 1996) and remain in this state for six to seven months. For
the population to grow, adult butterflies must survive through the winter. Inclement weather
conditions during winter can be catastrophic to these insects and have direct impacts on their
population size.
The abundance of host trees in coastal California and Central Mexico provide cover to
survive the cold of winter. The dense canopy of the host tree clusters in overwintering sites
provides protection from temperature fluctuations and rain, and thus, helps to conserve their lipid
storage (The fat energy reserve stored in their body from feeding on the nectar during migration),
and diapause state (Stenoien et al. 2018). The wet conditions during winter rain can lead to
freezing of the monarchs and can cause mortality as the water particles on these insects tend to
freeze them at just above freezing winter temperatures and forces the monarchs to burn their
energy reserves fast (Calvert et al. 1983; Anderson and Brower 1996). Therefore, dense tree
canopy and clusters provide shelter from rain and wind. This in-turn helps to conserve their lipid
storage as they rarely fly to find nectars during winters.
Monarch butterflies of both east and west of the Rockies start their migration during
spring from their overwintering sites and travel north towards the Southern Canadian border. In
spring, they begin their reproduction cycle and lay eggs on milkweed plants. The adult butterflies
and larva feed on the nectar and leaves of these plants (Malcolm et al. 1993). The availability of
milkweed plants during spring and summer is critical, as monarch butterflies lay eggs and feed
only on these plants. Weather changes in the spring and summer breeding grounds can affect the
plant phenology and thus change monarch migration patterns.

4
1.2 Western Monarch butterfly
The California coast hosts approximately 400 wintering sites for the western monarch
butterfly. Since 1990, there has been significant land cover changes in the winter sites in
California and in the overwintering sites in Mexico (Frey et al. 1992; Anderson and Brower
1996; Brower et al. 2016). The transformations are due to anthropogenic activities, and logging
of tree clusters. Land cover changes in winter sites in Mexico have been the subject of many
studies over the last 40 years, and it has been observed that the reduction in tree canopy cover
and below freezing temperatures, and precipitation can have a catastrophic effect on the monarch
survival in Mexico (Calvert et al. 1983). However, the changes in the overwintering habitat in
California has not been extensively studied. Since the early 1990’s the population of the western
monarch have been declining rapidly. Therefore, it is important to analyze changes over the last
two decades in the winter habitat of the western monarch butterflies.
The State of California also provides spring summer breeding habitat for the western
monarch population. During breeding season, monarchs depend on milkweed plants for survival
(Fig 2). The larva feeds on the milkweed leaves and enter pupal state on these plants. Therefore,
an abundance of milkweed plant is crucial to the monarch’s lifecycle. Milkweed plants are often
found in the cropland areas, open areas, and right of ways (Hartzler and Buhler 2000; Hartzler
2010; Wyrill and Burnside 1977; Pleasants and Oberhauser 2013).

5
Since 1990, increase application of the herbicide glyphosate in the cropland areas have
been a major cause of the decline of milkweed plants (Wyrill and Burnside 1977; Pleasants and
Oberhauser 2013; Pocius et al. 2017). According to the California Department of Food and
Agriculture, one-third of the country’s vegetables and two-thirds of country’s fruit and nut are
grown in California (California Agricultural Production Statistics 2018) . Application of certain
herbicide glyphosate can kill all vegetation in the applied area, and is used where genetically
modified herbicide-resistant crops are cultivated. The herbicide application kills the unwanted
plants in the croplands. Pesticide and herbicide application are a common practice on agricultural
Figure 2: Monarch Breeding and Milkweed Records of the Western Region

6
lands, potentially causing a decline in milkweed and thus affecting monarch survival (Wyrill and
Burnside 1977; Pleasants and Oberhauser 2013; Pocius et al. 2017).
1.3 Research Aim
This project aimed to quantify changes in land use and cover in the western monarch
butterfly winter habitat. Monterey county (Fig. 3) in California was selected as the study area.
Monterey county has 15 active overwintering sites out of the 400 known sites of the western
monarch butterflies. The county also has significant agricultural lands where herbicide
glyphosate is used on various crops each year. Monarch butterflies depend on the milkweed plant
for survival and these plants are known to be in agricultural areas, open areas and right of ways.
Increasing use of glyphosate is known to be causing decline in milkweed plants (Pleasants and
Oberhauser, 2013).
This project addressed the following spatial and temporal questions:
• What are the changes in land use in the habitat area? Can this change be correlated with
the change in monarch count?
• Which sections of the county are using glyphosate and how much is the pesticide use?
What is the trend at the county and section? (The county is subdivided and are
categorized into Meridian, Township, Range, and Section (MTRS). Section being the
small geographical area level)
• Are the locations of glyphosate application correlated spatially with the winter count of
monarchs

7
1.4 Study Area
Monarch butterflies arrive at overwintering sites in early November. The Monterey pine,
Eucalyptus, and Monterey cypress trees remain their home for the duration of winter. These
groves are located along the coast and at low elevation (Fig 3). Densely clustered trees provide
protection from wind, temperature fluctuations, precipitation, and direct sunlight. Throughout the
winter they move amongst trees within the cluster to find cover from wind and direct sunlight.
These sites act as micro-habitats. Due to their proximity to the ocean, these groves are
susceptible to change in wind direction and temperature change.
Figure 3: Study area Monterey County at township level and Monarch winter sites

8
California has a spring and summer migration path as well as breeding grounds. During
the start of spring migration, monarchs search for milkweed plants to feed on the leaves and
nectar. The state of California has the highest agricultural production in the US and depends
heavily on the use of herbicides and pesticides. Milkweed in the agricultural areas is declining
due to herbicide use (Stenoien, et al. 2018). Application of herbicide in cropland areas is
attributed to a major decline in the monarch butterfly population (Thogmartin 2017; Zipkin et al.
2012). For this project, application of herbicide glyphosate in the Monterey county from 1991 to
2015 showed, an average of 98,696 pounds of glyphosate usage with a minimum amount of
53,121 pounds used in 1995 and maximum of 161,587 pounds used in 2000. There has been a
steady increase in herbicide usage from 2003. Grapes (for wine and fruit consumption), and
uncultivated agricultural lands in the county used high amount of glyphosate every year.
Another factor affecting monarch populations are changes in the land use (Thogmartin et
al. 2017). California, and Monterey County in particular, is witnessing land cover changes due to
urbanization and wildfires. The land cover land change analyses in this project used NDVI
differencing method to quantify the land use change. The changes due to wildfire has not been
included in this study.
Extensive studies have been conducted on the eastern monarch population and their migration.
Their winter, spring, and summer migration habitat have been studied through statistical and
spatiotemporal analyses. Research and volunteer projects such as. Journey North and Monarch
Joint Venture are some the ongoing citizen participation projects. Through this project citizen
volunteers register their monarch sighting with spatiotemporal data. They tag the butterflies to
track the migratory path of eastern monarch butterflies.

9
In the west, Xerces Society conducts annual count of overwintering butterflies along the
California coast during Thanksgiving. From the literature review and search for related journal
shows, there are limited spatial analysis studies on the western monarchs and its habitat changes.
This is the biggest motivation for this project.
1.5 Previous Studies and Gaps
Western Monarch butterflies in California have been subject to few studies, insofar as
there has been little study of landcover cover change in the western monarch habitat. Some of the
gaps are listed in a report to the Commission for Environmental Cooperation 2009 (Davis et al.
2009). Per the report, there has been extensive loss of winter habitats in California with a decline
from 1990 to 1998. The habitat loss is due to tree loss, due to senescence or diseases, and urban
infrastructure development and other related human interference (Meade 1999, Leong et al.
2004) The eastern monarch population is monitored during winter, spring migration and
breeding season via citizen science project. Annual fourth of July count are conducted in the
monitoring sites, where volunteers gather egg, larva, pupal, and adult butterfly count. In addition,
land use land cover change, pesticide application affects have been studied as well (Zipkin et al.
2012; Pleasants and Oberhauser 2013; Batalden, Oberhauser, and Peterson 2007; Vidal, López‐
García, and Rendón‐Salinas 2014). Whereas, the western monarch population during spring and
migration is not monitored.
The first sighting of western monarch butterflies during the spring season is an indication
of their migration towards the Rocky Mountains. This journey is usually not well documented,
during spring and summer, and the population is not counted or surveyed, creating a void in the
migration data.

10
Much of the study of eastern monarch migration tags the butterfly, which documents the
location where the individual was found, indicating their migration path and route. Similar
documentation of the western monarch spring migration has not been conducted. Thanksgiving
count every year of the overwintering butterflies during the week of Thanksgiving is the only
record which provides information about the western population. This creates a huge gap in the
data. Spring, summer and fall migration data collection and documentation would help
understand the impact of land use and climatic changes on monarch populations and habitat.
Since spring and summer migration and population data is not readily available, changes
in monarch breeding habitat cannot be considered for this project. Therefore, the primary aim of
this project is to quantify changes in the winter habitats. Temperature datasets, herbicide
application timeline for overwinter sites, and breeding time were used as major data sets. The
data sets for this project were identified and collected based on studies conducted on eastern and
western monarch habitats. The next chapter on related work explains the studies and
observations made in the monarch winter and summer habitat, weather changes and, herbicide
application trend in both eastern and western monarch habitats

11
Chapter 2 Related Work
North American butterfly is known for its annual migration. During fall, they travel from
South of Canadian border to coast of California and forests in Central Mexico to the overwinter
sites, where they spend the winter months in the dense tree clusters (Brower et al. 2017; Brower
et al. 2002; Oberhauser et al. 2017). They start their migration up north every spring and start
their reproduction cycle by laying eggs on the milkweed plant. These host plants are crucial as
monarchs feed and lay eggs only on native milkweeds. Monarch life span during spring and
summer is around 2-6 weeks. It is the third or fourth generation which reach the south of
Canadian border by the end of spring summer migration. The generation of monarchs which
travel to over wintering sites in California and Mexico may survive up-to 6-9months
(Oberhauser 2004).
The multi-generation migration during spring-summer, and single generation fall
migration and overwintering survival is affected by its surrounding environmental changes in
terms of land use and weather changes (Brower et al. 2002; Pelton and Jepsen 2019; Espeset et
al. 2016).
The decline in the western monarch butterfly population is alarming (Dingle et al. 2005;
Pelton and Jepsen 2019). In 2018, only 28,429 butterflies were counted from 213 active sites
along the California Coast, (Xerces Society 2018). That is 85.2 % decline from the previous year
2017 and a 99.4 % decline since the 1980. Similar trend in monarch population has been
observed in the Monterey county, only 4000 butterflies were reported in 2018 in comparison
with the 2017 count of 35000 (Fig 4). The US Fish and Wildlife service states that the decline in
monarch population is due to loss of habitat for breeding, migration, and overwintering.

12
Monarchs are sensitive species, they need ambient weather conditions throughout winter
to survive. Changes in the habitat can affect the micro climate of the winter sites and can have a
direct effect on their winter survival. Thinning of tree clusters, can expose them to wind and
direct sunlight in winter. The inclement weather with precipitation and below freezing
temperature in winter habitat in Mexico in 2002 killed close to half a billion of monarchs
(Brower et.al 2002).
The decline in milkweed plants in spring summer habitat can affect their breeding cycle
and survival. Various study of the eastern monarch has observed and identified the changes in
winter and summer habitat and leading causes for such changes. In this chapter, all the previous
change observation of eastern monarch habitat has been extracted and the methodologies have
been documented.
Figure 4: Monarch Yearly Count, Monterey County. Source: Thanksgiving Count Xerces Society
R² = 0.0086
0
5000
10000
15000
20000
25000
30000
35000
40000
45000
50000
55000
60000
65000
70000
75000
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
Yearly Count
Year 1997 - 2018

13
2.1 Land use land cover changes
2.1.1. Overwintering sites
Eastern monarch butterfly roost on Oyamel fir trees in the forests of Central Mexico.
Oyamel fir forests create a microclimate suitable for monarchs to survive the winter. The thick
tree canopy protects monarchs from extreme winter temperature and precipitation, acting as a
blanket (Anderson and Brower 1996). The Oyamel trunks help regulate the temperature of the
monarchs when they are roosting on the trees. Degradation of these forests makes the monarchs
vulnerable to change in weather conditions and can lead to mass mortality during a winter storm
or due to temperature change (Brower 2002).
All overwintering sites in central Mexico were declared protected areas by the Mexican
federal government in 1980 (DOF 1980), established as a special conservation zones in 1986
(DOF 1986), and the government further expanded the protected area in 2000 (DOF 2000)
(Stenoien et al. 2018). Though these overwintering sites were declared protected areas,
deforestation and illegal logging have been reported and land cover changes have occurred
(Brower et al. 2002). In this analysis, stereo imagery of 42,020 ha area taken in 1971, 9184, 1999
were analyzed using GRASS and GIS. Based on the imagery, forest cover was qualified into 3
categories: conserved, semi-altered, and altered. From 1971to 1999, 44% of the forest area was
estimated to be degraded and fragmented, and the largest patch of the high-quality forest was
reduced from 27,115 ha to 527 ha, that means the high-quality forest cover is reduced to 21%.
From 1984 to 1999, some of the protected forest areas were degraded and fragmented at an
annual rate of 3% or more, in comparison to 1% from 1971 to 1984, indicating 31% reduction in
high quality forest in central Mexico. In this study, a subset area of 6596 ha inside the area
decreed as protected in 1986 was also conducted and showed 38% degradation. The study

14
indicates at this rate less than 10,000 ha of high-quality forest will remain in 20 years. Thick
forest covers and tree canopy provides shelter from rain and freezing temperature. As the winter
progresses, the monarchs shift to downward slope, therefore any fragmentation of forests makes
the monarchs vulnerable to survive.
Between 2001 and 2012, Vidal et al. (2013) conducted a GIS analysis examine trends in
deforestation and forest degradation of Monarch biosphere in Mexico using aerial and satellite
imagery and field surveys. They conducted a biennial analysis of aerial and satellite imagery to
determine changes in land cover due to large-scale illegal logging. To determine small-scale
logging and changes, 2001 image was compared with 2012 imagery. The 2007 SPOT
multispectral and panchromatic imagery were compared with 2005 and 2009 images. In addition
to satellite imagery, every year aerial imagery was acquired to determine the changes in land
cover. Results show 2179 ha forest in core zones were affected between 2001 to 2012. 1254 ha
were deforested, and 925 ha were degraded. Illegal logging persisted even after these forest areas
were declared as protected area. The study indicates though Monarch Butterfly Biosphere
Reserve was a protected area and there has been a decrease in large scale illegal logging since
2012, small scale logging has been clearing patches of crucial land and canopy inside the
Reserve could affect the future monarch population. Such GIS analysis is essential to monitor the
deforestation trend and visual maps aid the authorities in decision making.
Spatial modeling using GIS have been used to analyze the distribution pattern of
overwintering monarch butterflies. The area covering the maximum overwintering habitat was
delineated by Bojorquez et al. (2003). In this analysis, digital map layers, 149 overwintering
colonies, boundaries of 12 sanctuaries, DEM, landform, vegetation, and land use layers were
used to create a database. The Landsat TM was used to quantify land cover and vegetation data.

15
Weather and meteorological datasets were used to ascertain precipitation, the sun’s relative
position to analyze potential solar irradiation. Using GRASS, two spatial analysis approaches
were used to map potential overwintering habitat, which were: habitat suitability evaluation, and
logistic model using georeferenced records of overwintering colonies and GIS database. The
analysis results were used by 20 experts to generate potential overwintering habitat maps for
monarch butterfly biosphere reserve based on biological requirements of the butterfly. The
georeferenced records of the colonies revealed majority of the sites were at higher elevation
(2890 m and 91% sites), with low aridity, also with south -southwest aspect, with short distance
from streams, and on moderately steep slope. Multicriteria potential habitat analysis and experts’
opinion were thus combined to identify potential habitat areas to refine Monarch Butterfly
Biosphere Reserve
Landcover change from illegal logging and deforestation have been observed in the
overwintering sites in Mexico since 1977. However, the rapid decline in monarch population has
happened since 1990 (Stenoien et al. 2018). Land cover changes due to wildfires near the
monarch overwintering habitats have been observed (Vidal and Salinas 2014; Brower et al.
2017). Persistent drought conditions, low humidity, and high temperatures have resulted in
wildfires close to the overwintering sites in coastal California. Landcover change and
quantification for the western monarch butterfly habitat has not been analyzed. In this thesis
project land cover change due to drought could be seen in NDVI image difference analysis
results.
Quantifying the forest cover loss and land cover changes has been subjected to many
studies. Research indicates that there is ongoing illegal logging in the monarch winter habitat in
Mexico. The historical data from 1977 indicates that monarch form colonies at Sierra Chincua

16
arroyos in November and move to downslope areas by March as the winter progresses. The
downslope areas are where the logging have taken place. Scientists believe logging affect the
monarch remigration pattern with an assumption that monarchs might start their migration early.
Satellite imagery of Sierra Chincua before logging activity in 2013 was compared with 2015
imagery after logging. Recent study shows illegal logging of approximately 10 ha of forest in
Monarch butterfly biosphere reserve in 2015 (Brower et al. 2016). The satellite imagery of Sierra
Chincua before logging activity in 2013 were compared with 2015 imagery after logging. The
historical data from 1977 indicate that monarch form colonies at Sierra Chincua arroyos in
November and move to downslope areas by March as the winter progresses. The downslope
areas are where the logging have taken place. Scientists believe logging affect the monarch
remigration pattern with an assumption that monarchs might start their migration early.
In addition to habitat change analysis, monarch abundance has been measured in Mexico
since the 1994-1995 overwintering season and has been ongoing each year (World Wildlife
Fund). Areas occupied by the monarchs are combined and measured yearly. Western monarch
abundance in overwintering sites in California are measured through volunteer participation
annually during Thanksgiving under the supervision of Xerces Society. Since monarch roost in
small groves at each location, the actual count of monarch butterflies is recorded. The average
area occupied by monarchs until 2011 were 7.24 ha (average), with maximum of 20.97 ha during
1996-1997 and a low of 1.92 ha during 2009-2010 (Brower et al. 2012). Western monarch
abundance in the overwintering sites in California are measured through Volunteer participation
during Thanksgiving every year under the supervision of Xerces Society.
Deforestation due illegal logging is limited in the country and there are no reports of
illegal logging in the western monarch overwintering habitat. Deforestation leads to habitat

17
degradation and loss. However, other changes in land cover and land use can also lead to habitat
loss and degradation. Urban development, alterations of host tree cluster, tree trimming etc., have
all been observed near many of the winter sites in California (Pelton and Jespen 2019). Some of
the changes observed by Xerces Society were tree trimming, changes in tree cluster due to fallen
trees from disease and improper logging. Such change in the winter habitat is considered as land
cover change, as most of the winter sites are cluster of trees of the size 10 m
2
(approximately).
2.1.2. Spring and summer migration habitat
Land use land cover changes during the spring and summer migration of eastern
monarchs have been analyzed at multiple geographical regions and multi-temporal scales.
Studies of eastern monarch spring and summer migration habitat in the mid-west indicate the
loss of milkweed due to climate change, and due to use of glyphosate herbicide in agricultural
and open-lands (Pleasants and Oberhauser 2012; Pleasants 2015). Changing weather conditions
and frequent droughts, and their impact on have been affecting spring migrant butterfly breeding
cycle and reduce nectar availability for fall migrant butterfly (Pleasants and Oberhauser 2012;
Brower et al. 2006).
The stable hydrogen and carbon isotope ratios from 597 wintering monarchs from 13
wintering sites in Mexico were compared with the isotropic patterns measured from butterflies at
breeding sites over a migration cycle (Wassenaar and Hobson 1998). The study indicated that
monarchs at the over wintering sites in Mexico originated from the Midwest, in the United
States. 50% of the monarchs studied in the winter sites indicated that their origin was from the
breeding areas of Kansas, Nebraska, Iowa, Missouri, Wisconsin, Illinois, Michigan, Indiana, and

18
Ohio. These are the geographical areas where corn and soybean are produced extensively, and
the possibility of milkweed abundance in these areas.
Milkweed habitat is primarily found in the agricultural areas, roadsides and in the open
areas. Milkweeds are also found in the urban landscape. Studies indicate loss of milkweed
between 1999 and 2009, which corresponds to the time period when use of herbicide glyphosate
increased and increase in herbicide resistant corn and soybean production (Hartzler and Buhler
2000; Hartzler 2010; Oberhauser et al. 2001; Oberhauser et al. 2012). Oberhauser et al. (2012)
observed that monarchs foraged on milkweed found in corn and soybean fields. Land use
changes in agricultural areas is due to the use of herbicide-resistant corn and soybean crops. The
agricultural and nonagricultural area were mapped through citizen-science project and milkweed
data collected over a period of few years were analyzed to estimate the habitat loss. This study
was conducted from 1999 to 2010 and subsequent results showed an 81% decline in agricultural
milkweeds due to increased use of herbicide-resistant corn and soybean and glyphosate herbicide
application, and a 31% decline in nonagricultural milkweeds.
California has the highest agricultural production in terms of food related crops,
accounting to 13 percent of the nation’s total agricultural value (California Agriculture Statistics
Review, 2018). The leading commodities include livestock (dairy), grapes, almonds, berries, and
other nuts and vegetables. Production of corn and soybean grown in California is very low in
comparison with Iowa and other Mid-western States. Grapes and uncultivated agricultural areas
were found to use high amount of glyphosate in California, and particularly in Monterey county,
study area for this project. The use of herbicide glyphosate is increasing steadily since 1996, in
the Monterey county. However, there is no data to understand the changes in milkweed
abundance due to increase in glyphosate use.

19
Land use and land cover change due to development, wildfires, and drought are other
contributing factors responsible for western monarch decline. Recent data on western monarchs
indicate that after breaking up from the overwintering sites in spring, the monarchs travel only
shorter distance (According to USDA and Monarch Joint Venture 30-100 miles in a day). There
is no determined direction or path monarchs fly through. Once on their migration path, the
monarch butterflies lay their eggs on emerging milkweed. In a study conducted by Dingle et al.
(2005) used monarch data collections from museums and databases which stored monarch’s
sighting site’s geography and topology. DEMs, temperature, and a GIS layer of river locations
were used to analyze the spatial movement of the western monarch. The results indicate an
eastward and northward migration pattern, and monthly plots of data collection also show
monarch sightings in central and inland California from April to October, which indicates the
study area of Monterey county is also in spring-summer migration path of the monarchs
travelling from southern California coast and Mexico The study indicated that milkweed plants
are abundant in central and inland areas and analysis of land use and land cover change in these
areas are crucial for the conservation of host plant species.
2.2 Weather changes
Monarch butterflies are sensitive to weather and climatic changes. Above normal
temperature in winter can affect their lipid storage and diapause state which are crucial for winter
survival (Stenoien et al. 2018). Monarchs have a narrow micro-climate range to survive in the
winter. Freezing temperatures and precipitation can lead to mass mortality (Calvert et al. 1983).
Similarly, temperature fluctuations and weather change during spring and summer migration
have an effect on monarch and milkweed plants. Cooler temperatures and increased precipitation

20
can lead to longer developmental phase, above normal temperature, and drought-like conditions,
which reduce the lifespan of adults and larvae (Zalucki 1981 & 1982).
Studies have been conducted to understand the impact of climatic changes on the eastern
monarch butterfly overwintering habitat in Mexico (Oberhauser and Peterson 2003). Weather
data such as temperature, precipitation, and vapor pressure, and topographic data such as
elevation, slope and aspect were used with AVHRR NDVI at 1 km spatial resolution. These data
sets were used to analyze the climatic changes and develop a prediction model to forecast the
future of this overwintering habitat. Similar studies have been conducted in the southern
California overwintering habitats (Giffiths and Villablanca 2015).
Weather change and its implications during summer migration of the eastern monarch
population have been studied using ecological niche modeling by Batalden et al. (2007). In this
study, the geo-locations of the monarch eggs collected through the Monarch Larva Monitoring
Project (MLMP) were used. Descriptive stations of temperature datasets (max, min, mean
monthly, monthly solar radiation and monthly precipitation) and topography (elevation, slope,
aspect and tendency to pool water) were used for ecological assessments. Ecological Niche
Modeling has been used to identify areas adequate for overwintering for current and future
climate scenarios for the eastern monarch and the overwintering sites in Mexico (Oberhauser and
Peterson 2003). The study suggested that during breeding season, monarchs need to have an
optimum temperature and precipitation tracking northward with seasonal changes.
Study of climate change and its implications on monarchs and milkweed plants is limited.
Data on the summer breeding activity and location of the western monarchs are scarce as well.
Potential natal grounds of the western monarch have been analyzed based on milkweed
availability during summer and regional climatic conditions using “bottom-up” regression

21
models by Stevens and Frey. (2010). The study analyzed the relationship between moisture
variations at natal habitat and inter-annual monarch abundance variation at overwintering sites
between 1998 and 2007. Milkweed data at the county level, weather datasets, and monarch
abundance datasets were used for this study. Historical temperature data from the National
climatic Data Center (NCDC) was used to identify the regions in west that had degree days,
which could support late summer monarch recruitment. Presence and absence of milkweed
available for late monarchs was created as well. Natal ground maps for migrant generation of
monarchs was delineated as NCDC climatic divisions. The NCDC maps of western state climatic
divisions with super-imposed county-level boundaries were used to transform county wide
milkweed pattern to 54 climate divisions. For each of the 54 divisions degree-days available for
egg-to-egg monarch development were developed considering the temperature regimes averaged
across 30-year period from 1970–2000. In phase II of the study, NCDC data from 1998-2007
were used to estimate the amount of moisture available to regional flora each month. Monthly
Palmer Drought Severity Index values across the years 1998 to 2007 from NCDC climate
archives to estimate inter-annual variation of western moisture conditions in 54 divisions. Using
series of linear regression 29 divisions out of 54 divisions yielded favorable conditions for
monarch lifecycle development.
Various predictive models and strategies have been developed to identify ‘bioclimate
envelop’ of the species using climate variables (Pearson and Dawson 2003). According to
Pearson and Dawson, Bioclimatic envelopes are defined as the climatic component of the
fundamental ecological niche. Identifying the bioclimate envelop using correlation between
species distribution and climate variables could help in predicting the impact of climate change
on the species physiology. The bioclimatic modeling is based similar to ecological niche theory;

22
however, it uses only climatic variables and does not include environmental factors (Pearson and
Dawson 2003). A simple bioclimate model can provide a useful starting point when applied to a
suitable species at appropriate spatial scale. They can be used to identify the future changes to
the species and habitat distribution and assess the risk from climate change.
The changing climatic conditions can alter wind direction, and wind currents might
increase an insect’s energy demands and spatial disorientation (Liechti 2006). Migratory
behavior depends on the climatic variations as the timely arrival to the breeding grounds is
essential for the successful reproduction of the species. The growth and abundance of milkweed
plants and the onset of breeding of monarchs must align temporally for the species’ survival. The
temperature cues used by these insects may vary in comparison to the temperature cues used by
milkweed plants, and this mismatch and could potentially affect the breeding success of the
dependent species (Visser and Both 2005). Analyzing the trends in temperature changes along
the overwintering sites and spring-summer habitat could indicate if there is a mismatch in
milkweed growth period and monarch arrival during to the breeding grounds.
California hosts a good number of monarchs every year in its winter and breeding
grounds, nevertheless, assessment of western monarch habitat has been limited as monarchs of
the west are little less known (Stevens and Frey 2004). The literature in this chapter provided
inputs about techniques to assess the changes in habitat, data types, data sources and the
outcome. Comparing the studies conducted in winter habitat in Mexico to western monarch
winter habitat, this project identified that: quantification of land use land cover changes in the
western monarch winter and breeding habitat; trend in herbicide glyphosate application in
breeding habitat such as amount used, type of crops, and application year or time; and
identifying inclement weather pattern during over wintering season, are some of the gaps which

23
are taken up as topics of research interest in this project. In the next chapter, data types and their
sources, methodologies followed for analyses and the technical challenges faced in this project
are discussed extensively.

24
Chapter 3 Data and Methods
This project involved quantifying the land use land cover changes of the Monterey county from
1994 to 2018 at four-year interval, trend analysis of herbicide glyphosate use at section level (using
yearly data from 1990-2015), identifying inclement weather conditions using daily weather data, and
consolidating monarch yearly count data at each sites using WMTC data from 1997-2015.NDVI
differencing method was used to quantify the land use land cover changes using Landsat ARD
datasets; herbicide database was created and queried to analyze the trend in glyphosate application and
similarly, daily weather data was queried to identify inclement weather during overwintering.
3.1 Data and Methods Overview
This project used spatial and non-spatial data sets to examine land use land cover change
analysis. The methodology was based on that established by Lo and Shipman (1990) wherein
aerial photography, and topographical and geological datasets were used to quantify the land use
changes. There were five stages for quantifying land use, which were data extraction, data
capture, data integration, data analysis, and output of results.
As a preliminary step, data sets required for this project were identified (see Table 1).
This table lists all the data sources and type as used in this project. Some of the data listed in the
Table 1 were analyzed and were found to be less useful (marked in red) due to lower resolution
of the Landsat ARD imagery (30 meters resolution) in comparison with the size of the monarch
winter sites (few m
2
in area), therefore, were not used further. Though the DEM was identified as
a data source to quantify the winter sites terrain and the changes in the southward slope, where
monarchs roost during winter. However, western monarch butterfly winter sites are small

25
clusters of Monterey pine and eucalyptus trees (Espeset et al. 2016; Xerces Society Western
Monarch Overwintering Sites Database, 2017).
Data
Data
type
Data source
Data
format
Temporal
resolution
Spatial scale
Monarch
winter sites
Spatial Xerces Society Shapefile
Single year dataset,
updated 2018
Monterey
County
County
boundary
Spatial
California Department
of Pesticide Regulation
Shapefile Single year dataset
Monterey
County
Weather data
Non-
Spatial
University of
California Integrated
Pest Management
Tab
delimited
file
1990 - 2018
Monterey
County
Landsat ARD
imagery
Spatial USGS Raster
1994, 1998, 2002,
2006, 2010, 2014,
and 2018
Clipped to
Monterey
county
Monarch
winter count
Non-
Spatial
Xerces Society
Thanksgiving count
Tab
delimited
text file
1997 to 2018
Yearly count
Monterey
County
County
Township and
Section
Spatial
California Department
of Pesticide Regulation
Shapefile
Single year dataset
2018
Monterey
County
Pesticide Use
Report
Non-
Spatial
California Department
of Pesticide Regulation
at Section level
Tab
delimited
text file
1990 to 2015
(98578 records)
Monterey
County

3.2 Research Design
This project aimed to understand the usefulness of remote sensing and GIS in quantifying
changes in the monarch overwintering and breeding habitat. Creating a geodatabase, building
land use classes through Landsat imagery NDVI, creating spatial layer using the Pesticide Use
Reports (PUR), extracting the monarch sites in the study area, and spatial joining the weather
data from the nearest weather station to the were the first few steps of the analyses.
The research followed the Lo and Shipman (1990) land use change methodologies
following the five-step process of: data extraction, data capture, data integration, data analysis,
and output results and maps, for this project, as the project involved multiple data types and
different methodologies. The steps are explained in the following sections.
Table 1: Data, Type and Source

26
The Monterey county was selected as the study area for two reasons: (a) the county has 15
active over wintering sites, and (b) the county also has significant agricultural areas, making the
pesticide application analysis meaningful. All the analyses were carried out within the whole
county.
3.3 Data Description and Methods
This project aimed to understand the environmental and anthropogenic factors that has been
affecting the monarch butterfly population based on the various studies and observations
(Pleasants and Oberhauser 2013; Pocius et al. 2017; Oberhauser et al. 2001; Brower et al. 2012).
These studies indicate that changes in environment due to human activity around monarch winter
sites have affected the quality of monarch habitat. The changes in monarch butterfly
overwintering habitat has been analyzed through spatiotemporal analysis using spatial and
nonspatial data sets (Section 3.1, Table 1). Different methodologies were followed based on the
dataset selected for this project and were divided into land use land cover change analyses,
Pesticide use trend analyses, identifying weather conditions during overwintering and analyzing
the pattern of monarch count each year. The following sub sections explains the datasets, data
preparation and analyses methods followed.
3.3.1. Land use land cover changes
Application of remote sensing using satellite imagery is best suited for long-term habitat
change analysis, as in-situ monitoring of the host plant species and habitat is challenging as the
host plants and the insects are dispersed through the landscape.
Land use land cover change is one of the major causes of habitat loss. It leads to change
in biodiversity and sometimes pushes the sensitive flora and fauna towards extinction (Pelton et

27
al. 2019, Schultz et al., 2017). Land cover and land use changes are due to natural alterations and
human interventions. Human interactions with nature and introduction of chemicals are one of
the primary causes of habitat change and habitat loss (Falcucci et al. 2007). The habitat loss leads
to biodiversity hotspots, which needs to be identified and addressed in a timely fashion to stop
the species from extinction. Understanding past and current changes of land use helps analyze
the effect on ecology land degradation over time. Since monarch butterflies are migratory species
and travel long distances through different geographical regions and seasons, monitoring changes
via field observations may lead to inconsistencies and gap in the data. Satellite and aerial
imagery, on the other hand, is very useful in multi temporal analysis (Mouat et al. 1993).
Selecting satellite imagery depends on the size of the study area and its temporal
coverage. Initially, Landsat ARD (Analysis Ready Data) images of each year from 1991 to 2018
was considered for land use land cover change analyses. Four ARD tiles (Fig 5) covered the
Monterey county, and to maintain the temporal consistency ARD images between November 1
st

– December 31
st
of each year were included as search criteria. Due to unknown reasons all four
tiles for one single date were not available for some years, for example: one (or two) tiles out of
four was missing in the data, or, there was more than 10% cloud cover, or, there were Landsat
scan line error in the tiles. Since 2003, some of the images have been affected by scan line error,
thereby causing data gaps. Data rectification of scan line error using algorithms is a time-
consuming process. data was carefully selected, without any cloud cover or scan line error
images have been avoided, except 2010 image where H01V10 image tile has scan line error.
Landsat ARD (Analysis Ready Data) images of the Monterey county of the year1994, 1998,
2002, 2006, 2010, 2014 and 2018 were selected for quantifying the land use land cover change.

28
Landsat ARD products are processed from Landsat Collection 1 Level 1 scenes and are
ready to use for monitoring or assessing land use change. The Landsat ARD consists of top of
atmosphere reflectance, top of atmosphere brightness temperature, surface reflectance, surface
temperature, and quality assessment data along with metadata file. These ARD products are
available for conterminous US, Alaska and Hawaii through the following Collection 1 Level 1
products:
• Landsat 8 Operational Land Imager (OLI)/ Thermal Infrared Sensor (TIRS) Tier 1, Tier 2
• Landsat 7 Enhanced Thematic Mapper Plus (ETM+) Tier 1
• Landsat 4-5 Thematic Mapper (TM) Tier 1
The Landsat ARD products follows Tile-Grid system, where, each image is available in
5000 x 5000 30m pixel tile (Figure 5). The tiles are the part of the grid system which covers the
conterminous US, Alaska, and Hawaii separately. The dimensions of Landsat ARD tile pixels are
smaller than Landsat 4-7 (approx. 6,166 x 5,666 pixels), and Landsat 8 (approx. 6300 x 6000
pixels) and are identified with horizontal (H) and vertical (V) tile numbers. The highlighted area,
H01V09, H02V09, H01V10, and H02V10 in Figure 5 shows the tile coverage for the study area.

29
The Landsat ARD Surface Reflectance (Near Infra-Red (NIR) and Red bands) are used
calculate NDVI. In this project the land use change was analyzed by calculating the NDVI and
quantifying the difference in NDVI between the two temporal images. Surface reflectance band
specification is as mentioned in Fig 6 and 7. The surface reflectance bands 4 (Near Infra-Red)
Figure 5: Landsat ARD Tile and Grid format. Source: USGS Landsat ARD

30
and 3 (Infra-Red) of Landsat 4, 5 and 7, and band 5 (Near Infra-Red) and 4 (Infra-Red) of the
Landsat 8 were used to calculate NDVI.
Figure 6: Landsat 4-7 Surface Reflectance Specifications. Source: USGS Landsat.
Figure 6: Landsat 8 Surface Reflectance Specifications. Source: USGS Landsat

31
Initially, the NLCD (National Land Cover Data) of the study area was considered as the
primary data for the land use land cover change. The incompatibility of NLCD 1992 data and the
suitability issues of the Retrofit Land Cover data at the local level considered, however, Landsat
ARD was selected because of temporal resolution. Also, the NLCD1992 dataset has 21 land use
types, whereas NLCD from 2001, 2006, and 2011 have 16 land use types (classifications). This
difference in land use classes could lead to conflicts during land use change analyses. The NLCD
website warns about the difference in classes for land cover change analysis studies. Due to the
difference in the classification types of NLCD 1992 (21 classes) and NLCD 2001, 2004, 2011
(16 classes), NLCD dataset was not included in this project for land use change analysis as the
effects of different classes were unknown. Instead, calculating NDVI of the Landsat ARD and
creating land use classes based on the NDVI value was considered straightforward and the land
use change analysis was executed.
Landsat ARD data was downloaded from USGS Earth Explorer website and four grid
tiles covered the Monterey county (Fig 5). Initially, composite band image was created for each
tile using the surface reflectance bands. Composite Bands tool in Raster Processing ArcToolbox
creates a single raster dataset from multiple bands. This process was followed for all the four
tiles of Landsat ARD imagery for each year to create composite band imagery for each tile
(0109, 0110, 0209, and 0210). Then the four tiles of each year were mosaiced to form a single
image with composite bands (Fig 8). Mosaic to New Raster tool was used to create a new raster
dataset. A single image with composite bands was created for each year. The raster datasets were
then projected using Raster Project tool in ArcToolbox. This process was followed for each year
of the data selected and clipped to the county boundary.

32
NDVI Image Differencing
NDVI differencing method was selected for quantifying land use land cover change.
NDVI indicates greenness in vegetation and is useful in assessing plant health (Forkel et al.
2013). There are many suitable methods for assessing and modelling land use change such as
image differencing, image regression, and change vector analysis. Sometimes, a combination of
methods is used. In this project NDVI differencing and post classification using indices value has
Figure 7: Composite band image of Landsat ARD 1994

33
been used to quantify the land use land cover change. Near Infra-Red (NIR) and Red (R) bands
of the image are used to calculate NDVI as shown in the below (Equation. 1)

𝑁𝐷𝑉𝐼 =
(𝑁𝐼𝑅 − 𝑅 )
(𝑁𝐼𝑅 + 𝑅 )

The NDVI values range between −1 and +1(negative 1 to positive 1).The negative value
of the index approaching −1 indicate water, values close to −0.1 to 0.1 corresponds to barren
areas of rock, sand and built-up areas, values between 0.2 to 0.4 indicate moderate vegetation
and shrubs, and values above 0.5 to 1 indicate healthy vegetation and forest areas. A healthy crop
land also has higher NDVI values.
NDVI Range Category Map visualization colors
-1 to 0.1 Built up area and water body
0.1 - 0.2 Built-up and Open Area
0.2 – 0.35 Poor Vegetation
0.35- 0.628 Moderate Vegetation
0.628- 1.0 Healthy Vegetation
The Image Analysis and Raster Calculator tool in Arc GIS were used to generate the
NDVI images of 1994, 1998, 2001, 2006, 2010, 2014 and 2018. The Image Analysis tool uses
composite band image, and Infra-red and Red band must be manually selected prior to analysis.
Raster Calculator in Spatial Analyst can be used to perform NDVI using NIR and IR bands by
following Equation 1.
The NDVI value range from the analysis were divided into 5 categories based on the
index range. Every image of NDVI followed the same range and category for visualization and
Equation 1: NDVI, Where NIR is the Near Infra-Red and R is the Red band of the Landsat Imagery
Table 2:NDVI Pixel Range and Categories

34
further for NDVI image differencing. Reclassification tool was used to create land use
categories.
Discrepancies in NDVI pixel values were observed in 1998, 2006, 2010 and 2014
images. The NDVI value in some pixels were out of range (-1 to 1) with values ranging from
−1.1 to −10 on the negative side and values from 1.1 to 20 on the positive side as shown in
Figure 9. The images in the left are the NDVI maps with values ranging from -1 to +1, and the
images on the right indicate the pixels in the image which have out of range NDVI values, with
negative values in the range of -1.1 to -10 and positive values from 1.1 to 20. In the maps on the
right side, to identify the pixels having out of range NDVI value, they are represented in color
(blue, purple and pink), whereas, pixels within NDVI values of -1 to 1 are represented as ‘no
Figure 8: Out of Range NDVI values of 2002 and 2014 NDVI images

35
color’. The out of range NDVI values are not acceptable and their presence in the image affects
the median and mean values, and the resultant values during differencing of two NDVI images.
The Mask function in the Arc GIS was used to remove the unusual (out of range) pixel
values. Using the Mask function, a range of specific values can be mentioned and, any value
outside the minimum of maximum value of the range mentioned will be returned as No-Data.
Thus, the resulting NDVI maps had the values between −1 to 1.
To analyze the change from Year Y1 to Year Y2, image differencing technique was used
(Equation 2). NDVI image differencing indicates the change in the vegetation at pixel level
(30m) (SINGH 1989; Forkel et al. 2013; Ehsan and Kazem 2013).

∆𝑁𝐷𝑉𝐼 = 𝑁𝐷𝑉𝐼 𝑌 1
− 𝑁𝐷𝑉𝐼 𝑌 2

NDVI differencing involves subtracting two different date NDVI images. The pixel by
pixel subtraction of the NDVI image results in NDVI value ranging from −1 to +1, negative
value indicating positive change, values near zero indicate no change and positive values indicate
negative change . The resultant NDVI difference values were further reclassified into four
categories: Increase in healthy vegetation, increase in moderate vegetation, land cover
unchanged, and decrease in vegetation (loss). The reclassified categories in pixels were used to
represent percent change. Raster Calculator in Spatial analyst toolbox was used to subtract NDVI
images of two subsequent years (Fig 10).
Equation 2: NDVI Image Differencing

36
NDVI image differencing using the raster calculator was used to create a difference
image. For example, NDVI 1994 image was subtracted from 1998 image to identify the land use
change over the 4 years. Similarly, 1998 image was subtracted from 2002 image and so on.
3.3.2. Digital elevation model map
USGS National map 3D Elevation Program provides Digital elevation models at 30-
meter resolution. The elevation represents the topographic bare-earth surface. These maps are
used for hydrologic modeling and land use applications. In this project, the Landsat imagery was
overlaid on DEM to understand the topography of the habitat and detect the changes over the
temporal period.
During overwintering, monarchs need shelter from wind, direct sunlight, and
precipitation. Terrain plays an important role for monarchs in their selection of sheltering trees.
The eastern monarchs are regularly found roosting on trees on the south-west slope in their
Figure 9: NDVI image and NDVI differencing process

37
overwintering sites in Mexico (Oberhauser and Peterson 2003). Similarly, the western monarchs
prefer Monterey pine and eucalyptus tree cluster groves at a lower elevation, situated on south,
west or south-west slopes, for overwintering (Lane 1993; Frey et al. 1992). The slope, and
elevation of host trees in the overwintering sites along depends on the angle of the sun, the wind
direction, and the cluster of the surrounding trees (Leong 1990).
Monarch butterfly move between the cluster of trees when there is a change in wind
direction, precipitation or if there is direct sunlight, as they need protection to survive weather
changes during winter (Leong et al. 2004). However, the wintering sites in Monterey county are
very small in area comprising of few trees. The ground coverage of many sites could be less than
one pixel of Landsat image. Analyzing such microhabitat changes requires high resolution
topographical data, and satellite or aerial images. In this project, Landsat resolution could not
indicate the percent change in the tree density and landscape changes in the winter sites, but
changes at the county level was analyzed effectively. This study concentrated on changes in the
tree groves located in south and southwest direction and slope as the monarch butterfly prefer
such terrain. However, due to the size of sites and the image resolution, analysis could not
indicate any change.

38
3.3.3. Weather data
Weather change also play an important role in ecological niche of a species. Severe
weather change events can lead to decline of the species and unusual weather patterns and events
could lead to massive mortality of the species. Monarch butterfly is one such sensitive species.
50% of the monarch population can freeze at 24
0
F when their bodies are wet due to precipitation
and 100% mortality occurs at 18
0
F (Anderson and Brower 1996). Calvert et al. (1983) estimated
that around 2.5 million butterflies did not survive the 1980-1981 winter storm, followed by snow
fall and below freezing temperatures during the month of January. Another inclement event in
2002 reported to have killed nearly half a billion butterflies (Brower et al. 2004). It was therefore
imperative to know the unusual weather events during the last two decades (from 1990-2018) in
the wintering sites of western monarchs.
The daily weather data was collected in non-spatial text format from the nearest weather station
to the study area. These weather stations are monitored by University of California statewide
Integrated Pest Management (UC IPM). UC IPM weather data includes maximum and minimum
temperature values, precipitation, wind speed, wind direction, and solar radiation (Fig 7). The
Figure 10: Sample of daily weather data of the Monterey county from 1997-2016
Station Date_record Month Precip Temp_max Temp_min Wind_dir Wind_speed Solar_rad ETo Humidity_max Humidity_min
Monterey 1-Jan-97 1 0.29 67 61 SE 10 97 0.03 92.8 83.7
Monterey 2-Jan-97 1 0.55 65 57 SE 9 75 0.02 96.1 85.6
Monterey 3-Jan-97 1 0.01 61 44 NW 6 281 0.05 94.7 52.7
Monterey 4-Jan-97 1 0.03 51 39 S 3 97 0.02 93.7 71.7
Monterey 5-Jan-97 1 0.12 55 43 NW 7 290 0.05 95.7 58.2
Monterey 6-Jan-97 1 0 57 35 N 6 301 0.08 77.2 35.3
Monterey 7-Jan-97 1 0 59 31 S 4 275 0.05 90.8 42.6
Monterey 8-Jan-97 1 0 61 33 S 4 266 0.05 93.8 50
Monterey 9-Jan-97 1 0 62 34 S 4 281 0.05 96.3 50.5
Monterey 10-Jan-97 1 0 61 35 NW 4 273 0.05 97.2 60.6
Monterey 11-Jan-97 1 0.01 52 39 S 4 61 0 98.1 82.6
Monterey 12-Jan-97 1 0.27 55 44 SE 4 156 0.03 95.7 82.5
Monterey 13-Jan-97 1 0 51 39 NW 4 235 0.05 84.7 36.6
Monterey 14-Jan-97 1 0.28 47 34 SE 7 116 0.03 95.4 67
Monterey 15-Jan-97 1 0.15 63 45 S 4 215 0.04 96.9 66.7
Monterey 16-Jan-97 1 0 59 42 S 4 218 0.04 97.8 75.3

39
weather data was exported to MySQL, and queries were built to identify the days with below
freezing temperature, days with below freezing temperature (24
0
F) and precipitation. Monarch
population is also affected by above normal winter temperatures. It can accelerate the spending
of their lipid reserves (Anderson and Brower 1996; Brower 2004). Weather data was queried to
identify days with above normal winter temperature.
Select Date_record, Precip, Temp_max, Temp_min from weather_monteryfinal
where Temp_min<= 33
order by Temp_min asc, Precip;
/* different temperature values for Temp_max and Temp_min was used in the query */

3.3.4. Pesticide Use Report
The state of California is one of largest agricultural producer in the country. One-third of
the country’s vegetable and two-third country’s fruit and nuts are grown in California. The
production of corn was in the 1993 crop year was 430,804 tons and in 2016 the corn production
was 13,490,064 tons, according to the California department of food and agriculture reports. As
the total production of corn and other crops increased from 1993 to 2016, the production of other
crops has increased over time, indicating increase in the rate of cultivation in California.
Milkweed plants are found in abundance in corn and soybean croplands apart from the
milkweeds found near the right of ways and in the open areas (Pleasants and Oberhauser 2013;
Pocius et al. 2017; Oberhauser et al. 2001, Hartzler and Buhler 2000).). Monarchs depend on
milkweed for nectar and plant leaves during spring and summer migration for survival and
breeding. Monarch butterflies lay their eggs only on milkweed plants. The abundance of this host
plant species is essential for the entire monarch lifecycle (Ljiljana et al. 2016; Pleasants and
Oberhauser 2013; Hartzler 2010; Flockhart et al. 2012). Since the cultivation of genetically

40
modified crops of corn and soybean increased, the use of weed killing pesticides and herbicides
has increased over time. One such herbicide which has been observed to decrease milkweed
abundance is the use of glyphosate. Glyphosate is known to kill all the plant species when it
applied. Therefore, it is used with the herbicide resistant crops like genetically modified corn,
soybean, and cotton (Pocius et al. 2017). This herbicide is used in the agricultural lands where
genetically modified corn, soybean and cotton, which are resistant to glyphosate are grown in
California.
The California State Department of Pesticide Regulation (DPR) started reporting the
pesticide use in 1990. Under this program, pesticide use is reported monthly to the County
agriculture commissioner, with the application information such as: date and location of
application, type and amount of pesticide used (in pounds), type of crop for the pesticide
application, plant acreage, identification (ID) of the site, operator (pesticide user) and method of
application (ground or aerial). The pesticide application report also includes application in parks,
golf course, range lands, and right of ways. The report includes geographic location where the
pesticide was used, at the section level and includes, township.
For this project, only herbicide glyphosate application data from 1991 to 2015 (98423)
records) was downloaded in the non-spatial format with spatial representation of section level
GIS layer, making spatial display and GIS analysis possible. The California DPR also provides
shapefiles of the sections and townships for each county, which helped to create a relationship
between the tabular data of the PUR and section and township shapefile.
Studies indicate that the use of glyphosate in the cropland areas in the mid-west is the
main cause for the decline of milkweed in the cropland areas (Hartzler and Buhler 2000; Hartzler
2010; Wyrill and Burnside 1977). Quantifying the trend and rate of application over the last two

41
decades (1990-2018) in the study area was executed through GIS methodologies and querying
techniques.
The purpose of the PUR was to analyze the agricultural crop types in the Monterey
county and the amount of pesticide used by each crop type and its frequency. The PUR data table
contains the crop type, time of application, type of pesticide used, amount of pesticide used, areal
or ground application type, and Section ID. Shapefiles of township and section, and PUR table
was used to create a geodatabase. A simple relationship was created between PUR to section and
township shapefiles using the function ‘Relate’, for visualization purposes. The sections where
the pesticide glyphosate is used every year are highlighted in the Figure 12.

42

Figure 11: Glyphosate application from 1991 - 2015

43
PUR table (Fig 13) was exported into MySQL where queries were built to analyze the
crop type, pounds of pesticide used by crop type and year-wise, and the sections where the
pesticide was applied. Some of the queries are listed below:
• Sections where glyphosate was used frequently and the crop type:
SELECT COMTRS, COUNT (COMTRS), SITE_NAME,
SUM(POUNDS_CHEMICAL_APPLIED) FROM monarch_db.pur
GROUP BY COMTRS, SITE_NAME
ORDER BY COMTRS, SITE_NAME;
/*COMTRS is Section number; SITE_NAME is Crop type */
• Pounds of glyphosate applied each year:
SELECT YEAR, SUM(POUNDS_CHEMICAL_APPLIED) FROM monarch_db.pur
GROUP BY YEAR;
• Crop-wise glyphosate applied each year
SELECT YEAR, SITE_NAME, SUM(POUNDS_CHEMICAL_APPLIED) FROM monarch_db
.pur91_05
GROUP BY YEAR, SITE_NAME;
The results and maps of the PUR analyses are discussed in chapter 4.
Figure 12: PUR Table in ArcMap

44
3.3.5. Over wintering sites and Yearly count
There are over 400 overwintering sites along the California coast and there are over 200
active sites are where monarchs have been observed more than once since 2010 (Xerces Society.
Monterey county hosts up to 15 active sites, where monarchs have been sighted annually. For
this project, overwintering sites in Monterey county were selected from spatial data of sites
provided by Xerces Society. The yearly Thanksgiving count for these sites was used to analyze
population trend over the years.
3.4 Output Results and Maps
The results from all the analyses mentioned section 3.3in are discussed in detail in
Chapter 4. The results of land use land cover change analyses used maps, table, and graph
options to display the results. Displaying the results in map format provide better visualization
tools. Queries used for assessing PUR and weather data are displayed in Chapter 4 in the relevant
sections. All results are displayed using maps, tables, and figures to understand the trend in the
land use change over from 1990 to 2018.

45
Chapter 4 Results
This chapter reviews the outcome and results of the land use land cover analysis using
NDVI differencing method, pesticide application trend, and weather change analyses. The
change in each land use category is quantified and displayed in map, table, and graph format.
The results of PUR analyses and querying method used are also displayed in map, table, graph
format and the queries used are listed in relevant section. The weather change analysis to identify
the inclement weather conditions are displayed in table format and queries used are listed in the
subsequent sections.
4.1 Land use land cover change analyses
To analyze and quantify the land use land cover change, NDVI image of all the years
were classified based on the suitable NDVI value range and further reclassified into 5 categories,
using the Reclassify tool. The reclassified image was merged based on the grid values. The
merged pixels were used to derive the land use land cover category in percentage.
Image analysis and Raster calculator tools in Arc GIS were used to generate NDVI
images of 1994, 1998, 2002, 2006, 2010, 2014 and 2018. Image analysis tool uses composite
band imagery and NIR and IR bands were selected manually. Raster Calculator from the Spatial
Analyst toolbox and Image Analysis tool generated similar NDVI images and the results showed
similar NDVI value. Both the tools generate NDVI raster image with pixels denoting vegetation
index values ranging from −1 to 1. The images were reclassified based on the pixel range (Table
3, below) and the land use land cover category was created in the raster table. The resultant maps
created showed different land use land cover category. Number of pixels in each category was

46
used to create percentage of each land use land cover type and the results were displayed in table
and graph format.

NDVI range Category
-1 to 0.1 Water and built-up
0.1 to 0.25 Built-up and Open area
0.25 to 0.35 Poor Vegetation
0.35 to 0.628 Moderate vegetation
0.628 to 1 Healthy Vegetation

4.1.1. NDVI image generation and results
The NDVI image showed different land use land cover patterns, reclassifying the image
based on the determined pixel range helped to categorize the land use land cover categories
uniformly for all images. Reclassification helped to aggregate all the pixels belonging in the
NDVI range selected into one category. Land use land cover percentage in the study area were
further derived from the merged pixels (Ehsan and Kazem 2013). Figure 14 to 17 show the
NDVI reclassified images from 1994, 1998, 2002, 2006, 2010, 2014 and 2018. The NDVI
images were classified into different category based on the pixel range mentioned in the Table 3.
The range was decided based on the visual interpretation of the composite image and the Landsat
NDVI value representing greenness of the vegetation.
The Land use land cover change was observed in 2002, with decrease in moderate vegetation and
healthy vegetation compared to 1998 and 1994. The decrease was more than 5% in both the
categories and open areas or barren lands increased by 10%. However, moderate vegetation and
healthy vegetation showed improvement in 2006, 2010, and 2014. In 2006, there was an increase
in area in the waterbody due to rainfall during the season.
Table 3: NDVI pixel range and land use land cover categories

47

Table 3: NDVI Class and Land Cover Percentage
LC_type Land Cover Type
Count NDVI Pixel Count
LC_% Land Cover Percentage
Chart 1: Land use land cover percentage.

48

Figure 13: NDVI Image Reclassified 1994 and 1998

49

Figure 14:NDVI Image Reclassified 2002 and 2006

50

Figure 15: NDVI Image Reclassified 2010 and 2014

51

4.1.2. NDVI image differencing and results
The NDVI differencing method involves subtracting NDVI(t-4) with NDVI(t). The
difference in the image shows how the land use land cover has changed from the previous year.
Raster Calculator from the Spatial Analyst toolbox in Arc GIS was used create the difference
Figure 16: NDVI Image Reclassified 2018

52
image. Difference images of 1994-1998, 1998-2002, 2002-2006, 2006-2010, 2010-2014 and
2014-2018 were created.
When two NDVI images are differenced, the NDVI values at each pixel is differenced
with the other image. The resultant NDVI values of land cover with no change remains to be
around zero. The gain in vegetation tends to have higher positive value and the loss or
degradation in vegetation have values below zero. The waterbodies in the NDVI difference
image have values in range of 0 to −1. If the water body has reduced in size, then the change in
the NDVI values depend on the latest land cover trend. If there is gain in vegetation growth, it is
shown to have positive value range, and if the area has changed to open barren land, then it
shown to have values close to zero.
After the image differencing of the NDVI images, the difference image was compared
with the two NDVI images used. Figure 16 shows the glimpse of NDVI image of 2010 (a) and
2014(b), and the difference image (c). Area under the blue circle in all three images can be
observed to understand the land use change. Some areas which were green in image (a) has lost
vegetation (degradation or loss) in the image (b). Therefore, the image (c) shows those areas as
vegetation loss (in red). Similarly, increase vegetation cover is indicated by green color in the
image (c).

53

(a)
(b)
(c)
Figure 17: NDVI Images of 2010 (a), 2014 (b) and NDVI difference image of 2010-2014 (c)

54
Land use land cover gain and loss has been fluctuating from 1994 to 2018. The
vegetation loss in agricultural areas from 2002 to 2006 and 2014 to 2018 have been prominent.
The vegetation gain could be seen from 2006 to 2010. Similarly, vegetation loss in open lands
next to California coast was observed from 1998 to 2002 and 2002 to 2016. Observing the
vegetation near the agricultural areas and open areas (forest lands) is important as it might
indicate the availability of the milkweed plants in these areas. Analysis results indicate that the
size of water bodies has been reducing since 2006. The vegetation around some the monarch
sites have remained unchanged. The surrounding areas of the monarch sites have shown
fluctuations in vegetation cover. Thus, indicating that monarch sites have not been affected
negatively from 1994 to 2018.
The NDVI difference images were reclassified based on the pixel range and with visual
verification with NDVI images of two years. The reclassified maps have been displayed in
Figure 19 to 21.

55

Figure 18: NDVI Difference Image

56

Figure 19: NDVI Difference Image

57

Figure 20: NDVI Difference Image

58
The Chart. 2 shows the change in land use land cover every four years. The raster table
from the reclassified image was used to derive the percent change in the land cover. To simplify
the land use land cover change, the reclassified image was divided into four categories: Increase
in healthy vegetation; Increase in moderate vegetation; Land cover unchanged and Decrease in
vegetation (or loss). From year 2002 to 2010 moderate vegetation cover increased in open areas,
where as the healthy vegetation decreased significantly. The land cover changes from 1994 to
1998, and 2010 to 2014 have the same trend.
4.2 Pesticide Use Report analysis
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
1994 - 1998 1998 - 2002 2002 - 2006 2006 - 2010 2010 - 2014 2014 - 2018
Percent Land Cover Change
Increase in Healthy Vegetation Increase in Moderate Vegetation
Land cover Unchanged Decrease in vegetation (loss)
Chart 2: Land use land cover change in percentage derived from the reclassified NDVI difference image

59
The purpose of this analysis was to find the geographical locations inside the county
where the pesticide glyphosate was used, amount applied each year, the types of crop grown in
the county, application method, and the trend of glyphosate use over the years. Each application
of glyphosate reports the Section ID where it was applied, chemical name of herbicide, pounds
and gallons applied, crop type (Site name), aerial or ground application method indicator, permit
number, etc.
Analysis of the PUR table showed that approximately 68 crop types use the herbicide
glyphosate in the county, almost each year (Table 5 below). The analysis indicated herbicide
glyphosate was used in agricultural, commercial or institutional areas, structural areas, right of
ways, agricultural research commodity crops, aquatic areas and other application through
fumigation, each year.
Studies indicate that glyphosate is used where herbicide resistant, genetically modified
crops such as corn, soybean, and cotton are grown. Therefore, identifying the crop types which
use glyphosate was pertinent as part of the herbicide analysis. In this analysis, no records of use
of glyphosate in corn (exception for year 2006), soybean and cotton crops were found. The
results inferred either these crops are not grown in the county or these crops did not use
glyphosate. Further investigation of California Department of agriculture and Monterey county
crop reports (Yearly) showed (corroborated) that these crop types are not cultivated in the county
from 1991 to 2015. However, glyphosate was applied on corn only once in 2006 in one section
of the county, and approximately 52 pounds of chemical was used.
The herbicide analysis showed huge amount of glyphosate was used on grapes,
uncultivated agricultural and non-agricultural lands, a total of 1451965, 568613 and 230805

60
pounds of glyphosate respectively, from 1991 to 2015. Each year, these areas use high amount of
glyphosate
Chart 4 below, shows the crop type and pounds of glyphosate used each year. This chart
gives an insight into different crop types and the usage of glyphosate. It also shows increase in
use of glyphosate over the year for crops like grapes, citrus, lemons, right of ways and
uncultivated agricultural areas. Application of glyphosate increased from 1996 to 2001 and from
2004 there has been a steady increase each year till 2015 (Chart. 4 and 5).

61
1…
1…
1…
2…
2…
2…
2…
0
10000
20000
30000
40000
50000
60000
70000
80000
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Chart 3: Crops and year wise pesticide glyphosate applied

62

The types of crops grown in Monterey county are listed in the Table 5. Grapes,
uncultivated agricultural areas and uncultivated non-agricultural areas are reported to use high
amounts of glyphosate each year. Table 2 shows the cumulative pesticide used from 1991 to
2015 by each crop type. These are the same areas which use
Site_Name Sum(Pounds_Chemical_Applied)
Grapes, Wine 1,451,965
Uncultivated Agricultural Areas 568,613
Uncultivated Non-Ag Areas 230,805
Lettuce, Head 43,763
N-Outdoor Grown Transplant 31,174
Lemon 27,078
Asparagus (Spears, Ferns, Etc.) 15,415
Rangeland 14,170
Walnut (English Walnut, Persian Walnut) 13,065
N-Outdr Grwn Cut Flwrs Or Greens 12,974
Grapes 11,248
Table 4: Crop Type and Pounds of Chemical Used
Chart 4: Pounds of Glyphosate applied each year

63
Approximately 98000 records in the PUR table were queried in MySQL and the results
are displayed in Chart 3 and 4 and Table 5. Some of the queries designed are listed below:
SELECT YEAR, SITE_NAME, SUM(POUNDS_CHEMICAL_APPLIED) FROM PUR
Group by YEAR, SITE_NAME;
Data for the Chart 3 was derived from this query
SELECT YEAR, SUM(POUNDS_CHEMICAL_APPLIED) FROM PUR
Group by Year
Order by SUM(POUNDS_CHEMICAL_APPLIED) desc;
Data for the Chart 4 was derived from this query

SELECT COMTRS, Count (COMTRS), SITE_NAME,
SUM(POUNDS_CHEMICAL_APPLIED) FROM PUR
GROUP BY COMTRS, SITE_NAME
ORDER BY COMTRS, SITE_NAME;
Data to derive the pesticide use at section level was derived from this query. The result
table was exported to Arc GIS where relationship between this table and section shapefile was
created for further querying and analysis.
The results from queries were exported as table to Arc GIS to understand the
geographical locations where glyphosate was applied. Initially, all the Section id listed in PUR
were selected and displayed on map (Figure 12, Chapter 3). Though some of the areas are close
to monarch winter sites, correlation could not be established with the monarch population
decline. Grapes, uncultivated agricultural and non-agricultural lands, which use high amount of
glyphosate is located geographically on the map in Figure. 22 and 23

64

Figure 21: Glyphosate application in sections and towns based on crop type

65

4.3 Weather change analysis
The daily weather data from nearest weather stations located inside the Monterey county,
managed by Institute of Pest Management, was used to identify the days of inclement weather in
the past, i.e. from 1991 to 2015. The weather data table was queried in MySQL and the intent
was to find the number of days during over wintering season which had either below freezing
temperature with precipitation or days with above normal winter temperature. The query looked
for days with below freezing temperature (24
0
F to 16
0
F) with precipitation, which could be
Figure 22: Glyphosate application in sections and towns based on crop type

66
catastrophic to monarch butterflies. Monarchs cannot survive when their wings are wet in
freezing temperature as the water droplets from ice crystals and reduces the butterfly’s body
temperature. The query also searched for temperatures of 60
0
F and above as it tends to exhaust
lipid reserves in the monarch body and forces them to come out of hibernation and look for food
source (nectar and fresh water source), thus, burning their energy reserve (Calvert et al. 1983;
Anderson and Brower 1996; Stenoien et al. 2018). Daily maximum and minimum temperature
values, and precipitation data from November 1
st
to March 31
st
of the following year were
queried.
The query designed for this analysis are follows:
SELECT Station, Date_record, Temp_max,Temp_min, Precip FROM
monarch_db.weather_montery
WHERE Month IN (1, 2, 3, 11, 12) /* from November to March of the following year*/
AND Temp_min <= 30
ORDER BY Temp_max asc, Temp_min asc, Precip desc;
This query returned 159 records where the minimum temperature values were below 30
0
F, between November to March of the following year. Only two dates Feb-18-2006 and Dec-14-
2008 recorded precipitation of 0.14mm and 0.12 mm respectively, with minimum temperature of
29
0
F.
Similar query where the maximum temperature above 70
0
F during winter season (Nov,
Dec, Jan and Feb) was executed. The query returned 455 records, each indicating days where the
temperature was above 70. The result was further queried to identify the days with highest
recorded temperature between November to February. The first two weeks of November 1997,
2005, 2007, 2008, 2009, 2010, and 2014, recorded temperature above 90
0
F (hart 5)

67
SELECT Date_record, Precip, Year, Temp_max, Temp_min, Month FROM
weathertmax70new
WHERE (Year, Month, Temp_max) in
(SELECT Year, Month, max(Temp_max) FROM weathertmax70new
GROUP BY Year, Month
ORDER BY Year, Month;

Temperature above 80
0
F and 90
0
F were observed during the first two week of
November when the monarch colonies are not fully formed and the rooster sites are not fully
established, and during the last two weeks of February, just before the onset of spring migration.
Though studies do not indicate the effect of high temperature days on monarch butterfly during,
this analysis is only an observation of the changing temperature pattern during overwintering
season. However, studies do indicate that higher temperatures could trigger early onset of spring
migration, and warmer temperatures were observed during February, but there is no data as to
when the monarchs start their spring migration from overwinter sites in California each year.
72
77
92
74
72 72
81
77
73
81
75 75
72
76
75 75
76
78 78
79
85
71
84 84
86
73
77
73 73
74
72
71
75 75
76
78
73
91
77
73 73
87
84
80
79
82
90
73
72
77
92 92
77
84
81
92
75
78
94
78 78
79 79 79
77
87
73 73
79
83
80
84
80
91
80
85
82
74
71
84
70
75
80
85
90
95
100
29 -Jan -97
16 -Feb -97
2 -Nov-97
30 -Dec -97
13 -Nov-98
25 -Nov-98
17 -Dec -98
3 -Jan -99
27 -Feb -99
1 -Nov-99
26-Dec-99
27 -Dec -99
19 -Jan -00
7 -Feb -00
3 -Nov-00
28 -Nov-00
5 -Dec -00
2 -Jan -01
4 -Jan -01
4 -Feb -01
8 -Nov-01
11 -Jan -02
22 -Feb -02
27 -Feb -02
20 -Nov-02
4 -Dec -02
31 -Jan -03
6-Nov-03
18 -Nov-03
18 -Dec -03
8 -Jan -04
12 -Feb -04
2 -Nov-04
17 -Nov-04
17 -Dec -04
19 -Jan -05
1 -Feb -05
18 -Nov-05
24 -Dec -05
13 -Jan -06
24 -Jan -06
8 -Feb -06
7 -Nov-06
7 -Dec -06
9-Jan -07
17 -Feb -07
3 -Nov-07
5 -Dec -07
1 -Jan -08
12 -Feb -08
16 -Nov-08
17 -Nov-08
6 -Dec -08
13 -Jan -09
2 -Feb -09
2 -Nov-09
4 -Jan -10
17 -Feb -10
3 -Nov-10
25 -Jan -11
12 -Feb -11
1-Nov-11
2 -Nov-11
12 -Jan -12
9 -Feb -12
5 -Nov-12
20 -Jan -13
21 -Jan -13
15 -Feb -13
1 -Nov-13
25 -Dec -13
17 -Jan -14
24 -Feb -14
8 -Nov-14
25 -Jan -15
14 -Feb -15
22 -Nov-15
3 -Dec -15
27-Jan -16
16 -Feb -16
Chart 5: Highest temperature recorded monthly since 1997 to 2016

68
This analyses and results were compared with the monarch yearly count (Chart 6) to understand
if warmer winter temperatures had any effect on the monarch population.
Monarch yearly count takes place during the week of Thanksgiving in the month of November.
This is the time period when monarch colonies and roosting sites are established for that winter
season. Comparing the Thanksgiving count and high temperature data of a specific year do not
indicate whether the decrease in population count from the previous year is due to the warmer
temperature recorded.
The weather change analyses results did not indicate any above normal temperature days
during winter (December January and February) . The historical weather data analyzed did not
show unusual weather patterns. Therefore, there no correlation between weather change and
monarch count during over wintering season could be established.
4.4 Conclusions
0
10000
20000
30000
40000
50000
60000
70000
80000
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
Monterey Monarch Count
Years
Chart 6: Monarch Yearly Count, Monterey County. Source: Xerces Society

69
This project aimed to quantify the land use land cover change, analyze the pesticide
usage trend, and to identify unusual weather patterns during the over wintering season. All the
analyses were successful. The NDVI was a good indicator of the vegetation health, especially in
the open areas and agricultural lands. The loss in healthy and moderate vegetation was visible in
2002 (Chart 2), however, there is no significant change in the land use land cover in the
successive years. No other type of land cover change was noticed from one year to the other. The
NDVI differencing of the two consecutive images showed the trend in vegetation gain, loss and
unchanged land use. The vegetation loss from 1998 to 2002 and 2006 to 2010 in open areas and
fluctuations in agricultural trend is evident through NDVI differencing. This method could be
used in areas where ground truthing measurement is difficult to collect and supervised
classification cannot be followed.
PUR analysis provided insights into the glyphosate application trend and types of crops
and the geographical locations where they are grown. Some of the sites were close to the
pesticide application areas (Fig 22 and 23, above). Grape cultivation is abundant in the county,
which indicates the changes of milkweed to be found in those lands are to be low. However,
there are many uncultivated agricultural and non-agricultural lands where the glyphosate use has
been known to be high. Without ground survey, it is difficult to assume the presence or absence
of milkweed in these areas.
Analyzing the weather data was straightforward, as it concentrated on identifying unusual
weather pattern. The analysis was satisfactory as the results did not show days with inclement
weather, which suggests there no catastrophic weather incidents in the past, which could have
been harmful for the western monarchs.

70
The discussion of results from this project are discussed in the next Chapter. The future
scope of research for better understanding the relationship between monarch population and land
use land cover change, selecting different data sets, analysis methodologies, modeling and results
expected are discussed in Chapter 5.

71
Chapter 5 Discussion
In this chapter, conclusions from the results of the analyses, and the future research that
could be informed by this project are discussed in sections below. The datasets used in this
project overall were useful in helping to understand the geographical locations of the monarch
butterfly wintering sites and the land use land cover patterns in Monterey county. Analyses at the
county level was imperative to understand the changes around the monarch winter sites. Though
the data selected was relevant for the analyses and yielded a desirable outcome, due to spatio-
temporal issues the results did not show change results at high resolution, specially near the
monarch sites. The following sections explain the results from each analysis and their
correlation.
5.1 Conclusion
The land use land cover change analysis shows that from 2006 to 2010 there was an
increase in healthy vegetation from 2% to 5%, the highest in comparison with all other years
examined in this study. The increase was primarily in the agricultural areas. The moderate
vegetation cover however was reduced by 30% between 2006 and 2010, as compared to 2002 to
2006. The moderate vegetation was present in open areas and forest lands. Between 2002 and
2006, and 2006 to 2010, PUR analysis indicates that glyphosate use increased after 2006 after a
sharp reduction of use in 2002. The monarch count increased in 2003 in comparison to 2002 and
previous years. In 2002, 11,593 butterflies were counted for and in 2003 the number grew to
68,970. In 2006 there were approximately 60,000 butterflies and in 2007 only 15,426 butterflies
were counted. Similarly, monarch population declined from 1997 to 2002 when the herbicide
application peaked during the same time period. This might indicate a reduction in milkweed

72
cover in the agricultural areas in Monterey county. This is a significant finding with an
assumption that the milkweed availability decreased when glyphosate application increased, and
it in turn was the cause of decline in monarch population. It suggests that the glyphosate
application in the previous year either killed or reduced the milkweed abundance and when the
monarchs started spring and summer migration the next year, there were few milkweeds nearby.
Thus, these analyses might indicate scarcity of milkweeds may have led to the decline in
monarch count the next year. Whenever there is a reduction in vegetation in agricultural areas
(Figure 20 & 21), there has been reduction in glyphosate use and increase in monarch count. This
calls for further analyze the land use land cover change and pesticide application trend of
geographical areas which are in the monarch butterfly spring-summer migration path.
The weather change analysis did not show any inclement weather pattern during winter,
thus indicating the weather has been conducive for the monarchs without any catastrophic
events. No correlation could be established between weather pattern and the monarch count as
the warmer temperature were recorded before the roosting sites and colonies were established,
and Thanksgiving count. Minimum temperatures below 30
0
F were noticed, however there were
no precipitation recorded, and the maximum temperature recorded were in the range of 45-65
0
F
for those days, which indicated these weather conditions were safe for the monarchs to survive.
This project could establish a relationship between the land use land cover change,
pesticide use, and effect on monarch count. However, the land use land cover changes along the
spring summer migration should be quantified to understand the cause leading to the decline in
monarch population. It should be noted that the land use land cover changes in the winter habitat
and spring-summer migration habitat will have an impact on the entire population of western
monarch butterfly count.

73
5.2 Future scope of work
There are many change detection techniques which can be applied to assess land use land
cover changes, forest or vegetation change, deforestation, landscape, and environmental changes.
Some of the change detection provides information about: change and rate of change of an area,
spatial distribution, and accuracy of change detection results.
For most change detection techniques, thresholds must be identified to differentiate
change from no change. Though algebraic methods such as image differencing, image
regression, NDVI, and image ratioing using thresholds are adaptive and commonly used, these
methods however cannot provide change matrices, and selecting a suitable threshold to identify
the change is challenging (Lu et al. 2004). Selecting a suitable threshold is either through: (i)
interactive procedure or manual trial and error by the analyst to reach the suitable threshold or,
(ii) statistical method by selecting a suitable standard deviation from a class mean. The
limitations of these methods are that the resultant difference might have an influence of
atmospheric effect or subjective approach.
For more accurate change detection analysis, different change detections can be
combined. The band pair differencing of spectral bands could be used with Principal Component
Analysis (PCA) for multi-spectral image difference along with the application of other
operations to combine different information from multi-spectral bands to a single image (Gong
1993; Lu et al. 2004). Post classification comparison is another method used for detecting land
use change which provides a change detection matrix. This technique is useful for ‘from-to’
change analysis when used with image differencing. However, post classification is of limited
use when used with historical data, as classification technique might not yield accurate results.

74
More advanced and frequently used models are: Generalized Linear Models (GLM) and
integrated GIS and Remote sensing (Weng 2002) The advantage of using GLM for land use
change detection is that it could be used in Species Distribution Modelling (SDM). SDM is
suitable for distinguishing the factors affecting the ecology of a species suing topo-climatic
predictors.
The datasets identified for this project could also be used for habitat suitability mapping
(HSM). The habitat suitability model is best utilized when there is significant data on the actual
observation of the species. For this project, we have the count of the monarch butterflies
observed every year from 1997 to the present to confirm map predictions, as well as an
understanding of habitat conditions. The variables used for HSM is like that of SDM, thus the
usefulness of this model for this project shall be considered as the future scope of work.
The datasets identified for this project could be potentially used for direct assessment and
quantifying the land cover changes in the monarch habitat and the surrounding areas. The change
detection and quantifying the change help in understanding the habitat fragmentation (if
incorporated in this thesis), which could potentially indicate the cause for absence of monarch
butterflies in some of the winter site. Exploring PCA, GLM, suitability of HSM and SDM for
identifying the absence of monarch butterflies or milkweed abundance in some of the significant
sites shall be undertaken as future scope of work.
Different datasets such as Cropland Data layer (CDL) from National Agricultural
Statistics Service, US Department of Agriculture (USDA), to quantify changes in the agricultural
areas and high resolution multi-spectral imagery and DEM could be explored to analyze the
changes in the monarch winter habitat sites. The Xerces Society for Invertebrate Conservation
has been monitoring the tree clusters in the monarch winter sites. High resolution multi-spectral,

75
multi-temporal imagery tiles covering the site location shall be explored for monitoring the
changes in the sites, especially in the southward slope areas, where monarchs prefer to roost
during winter.
In addition to capturing daily weather data and calculating the growing degree days
(GDD) would be the next step in understanding the impact of temperature on the lifecycle of the
monarch butterfly. The effects of weather can influence adult monarch activity or juvenile
growth, and the growth of milkweeds (Zipkin et al. 2012). It is also helpful in understating the
temperature during spring-summer as a form of climate envelop indicating the optimum
conditions for the adults and juvenile to thrive. Correlating GDD and PUR might indicate the
milkweed abundance and monarch survival capability.
The project focused on identifying days with inclement weather conditions during
overwintering months, such as days with below freezing temperature, precipitation and above
average winter temperatures. The data and results shall be further studied and queried to
understand the cumulative weather conditions, and the trend during overwintering months and
spring-summer months for each year. The trend maps then shall be compared with the previous
and next year data, to understand the effect of changes in weather conditions during spring-
summer and winter, on monarch population and vice-versa. Such analyses when combined with
land use lover changes might indicate the vegetation health and host plant quality, both crucial
for monarch butterfly to survive. Further analysis of the weather data is therefore crucial to
understand cause of monarch population decline.
In this project, herbicide glyphosate application analyzed the crops types and the
geographical areas where it was used, and the amount applied. The future work shall include
identifying all the agricultural areas in the county where glyphosate is used and overlap with

76
NDVI images and NDVI difference image to understand the vegetation type (such as: moderate,
healthy, open areas, etc.) and amount glyphosate used. Since glyphosate is known to kill all the
vegetation present (except herbicide resistant crops), it would be interesting to identify the time
period when the large amount of glyphosate is used and analyze the NDVI maps of before and
after the application, specially in open areas (uncultivated agricultural areas and non-agricultural
areas). As the future scope of the project, this analysis is being executed.
This concludes the project work, the analyses executed, results, and future scope of work.
This project helped to understand the main factors affecting the monarch population. The land
use land cover change pattern overs the years showed a relation with application of glyphosate,
which affected the monarch yearly count. As mentioned in the future scope of work, a
combination of change detection methods and datasets could be explored to find a method which
could applied to the whole of western monarch spring-summer and winter habitat.

77
REFERENCES
Anderson, J. B. and L. P. Brower. 1996. "Freeze-Protection of Overwintering Monarch
Butterflies in Mexico: Critical Role of the Forest as a Blanket and an Umbrella."
Ecological Entomology 21 (2): 107-116. doi:10.1111/j.1365-2311.1996.tb01177.x.
Ahlqvist, Ola. 2008. "Extending Post-Classification Change Detection using Semantic Similarity
Metrics to Overcome Class Heterogeneity: A Study of 1992 and 2001 U.S. National Land
Cover Database Changes." Remote Sensing of Environment 112 (3): 1226-1241.
doi:10.1016/j.rse.2007.08.012. https://www-sciencedirect-
com.libproxy1.usc.edu/science/article/pii/S0034425707004026.
Batalden, Rebecca V., Karen Oberhauser, and A. Townsend Peterson. 2007. "Ecological Niches
in Sequential Generations of Eastern North American Monarch Butterflies (Lepidoptera:
Danaidae): The Ecology of Migration and Likely Climate Change Implications."
Environmental Entomology 36 (6): 1365-1373. doi:ENISGO]2.0.CO;2.
https://academic.oup.com/ee/article/36/6/1365/502715.
Bojórquez‐Tapia, Luis A., Lincoln P. Brower, Guillermo Castilleja, Salvador Sánchez‐Colón,
Mariano Hernández, William Calvert, Salomón Díaz, Paola Gómez‐Priego, Georgina
Alcantar, and Erika Daniela Melgarejo. 2003. "Mapping Expert Knowledge: Redesigning
the Monarch Butterfly Biosphere Reserve." Conservation Biology 17 (2): 367-379.
https://s3.amazonaws.com/academia.edu.documents/39162611/00b7d53c9679e2cfb80000
00.pdf?AWSAccessKeyId=AKIAIWOWYYGZ2Y53UL3A&Expires=1537778203&Sign
ature=ak52fX49I21jHHSLpbpBiAelIl0%3D&response-content-
disposition=inline%3B%20filename%3DMapping_Expert_Knowledge_Redesigning_the.p
df.
Brower, L. P., D. R. Kust, E. R. Salinas, E. García-Serrano, K. R. Kust, J. Miller, C. Fernandez
del Rey, and K. Pape. 2004. Catastrophic winter storm mortality of monarch butterflies in
Mexico during January 2002. In the Monarch Butterfly: Biology and Conservation, edited
by K. S. Oberhauser and M. J. Solensky, 151–166. Ithaca: Cornell University Press.
Brower, L. P., Taylor, O. R., Williams, E. H., Slayback, D. A., Zubieta, R. R. And Ramírez, M.
I. (2012), Decline of monarch butterflies overwintering in Mexico: is the migratory
phenomenon at risk? Insect Conservation and Diversity, 5: 95-100. doi:10.1111/j.1752-
4598.2011.00142.x
Brower, Lincoln P., Daniel A. Slayback, Pablo Jaramillo-López, Isabel Ramirez, Karen S.
Oberhauser, Ernest H. Williams, and Linda S. Fink. 2016. "Illegal Logging of 10 Hectares
of Forest in the Sierra Chincua Monarch Butterfly Overwintering Area in Mexico."

78
American Entomologist 62 (2): 92-97. doi://doi.org/10.1093/ae/tmw040.
https://academic.oup.com/ae/article/62/2/92/1751792.
Brower, L, Williams, E, Fink, L, Slayback, Da, Ramirez, Mi, Van Limon Garcia, M, Zubieta, R,
et al. 2011.“Overwintering Clusters of The Monarch Butterfly Coincide With The Least
Hazardous Vertical Temperatures In The Oyamel Forest.” Journal of the Lepidopterists
Society 65, no. 1 (April 20, 2011): 27–46. http://search.proquest.com/docview/867742853/.
Calvert, William H., Willow Zuchowski, and Lincoln P. Brower. 1983. "The Effect of Rain,
Snow and Freezing Temperatures on Overwintering Monarch Butterflies in Mexico."
Biotropica 15 (1): 42-47. doi:10.2307/2387997. https://www.jstor.org/stable/2387997.
Caswell, Hal. 2000. "Prospective and Retrospective Perturbation Analyses: Their Roles in
Conservation Biology." Ecology 81 (3): 619-627. doi:PARPAT]2.0.CO;2.
https://esajournals.onlinelibrary.wiley.comdoiabs10.1890/0012-
9658(2000)081%5B0619:PARPAT%5D2.0.CO%3B2.
Choudhary, Surendra Singh, P. K. Garg, and S. K. Ghosh. 2013. "Drought Analysis Using
Digital Image Processing & Meteorological Data." International Journal of Advanced
Remote Sensing and GIS 2, no. 1 (2013): pp-280. ISSN 2320 -0243. http://technical.cloud-
journals.com/index.php/IJARSG/article/viewFile/Tech-154/pdf
Dingle, H., M. P. Zalucki, W. A. Rochester, and T. Armijo-Prewitt. 2005. "Distribution of the
Monarch Butterfly, Danaus Plexippus (L.)(Lepidoptera: Nymphalidae), in Western North
America." Biological Journal of the Linnean Society 85 (4): 491-500.
doi://doi.org/10.1111/j.1095-8312.2005.00512.x.
https://academic.oup.com/biolinnean/article/85/4/491/2691594.
D T Tyler Flockhart, Tara G Martin, and D Ryan Norris. 2012. "Experimental Examination of
Intraspecific Density-Dependent Competition during the Breeding Period in Monarch
Butterflies (Danaus Plexippus)." PLoS One 7 (9): e45080.
doi:10.1371/journal.pone.0045080. https://www.ncbi.nlm.nih.gov/pubmed/22984614.
Ehsan, Sahebjalal and Dashtekian Kazem. 2013. "Analysis of Land use-Land Covers Changes
using Normalized Difference Vegetation Index (NDVI) Differencing and Classification
Methods." African Journal of Agricultural Research 8 (37): 4614-4622.
doi:10.5897/AJAR11.1825.
Espeset, Anne E., Joshua G. Harrison, Arthur M. Shapiro, Chris C. Nice, James H. Thorne,
David P. Waetjen, James A. Fordyce, and Matthew L. Forister. 2016. "Understanding a
Migratory Species in a Changing World: Climatic Effects and Demographic Declines in
the Western Monarch Revealed by Four Decades of Intensive Monitoring." Oecologia 181

79
(3): 819-830. doi://doi.org/10.1007/s00442-016-3600-y.
https://link.springer.com/article/10.1007/s00442-016-3600-y#citeas.
Faldyn, Matthew J., Mark D. Hunter, and Bret D. Elderd. 2018. "Climate Change and an
Invasive, Tropical Milkweed: An Ecological Trap for Monarch Butterflies." Ecology 99
(5): 1031-1038. doi://doi.org/10.1002/ecy.2198.
https://esajournals.onlinelibrary.wiley.com/doi/abs/10.1002/ecy.2198.
Flockhart, DT Tyler, Jean‐Baptiste Pichancourt, D. Ryan Norris, and Tara G. Martin. 2015.
"Unravelling the Annual Cycle in a Migratory Animal: Breeding‐season Habitat Loss
Drives Population Declines of Monarch Butterflies." Journal of Animal Ecology 84 (1):
155-165. doi://doi.org/10.1111/1365-2656.12253.
https://besjournals.onlinelibrary.wiley.com/doi/abs/10.1111/1365-2656.12253.
Forkel, M., N. Carvalhais, J. Verbesselt, M. D. Mahecha, C. Neigh, and M. Reichstein. 2013.
"Trend Change Detection in NDVI Time Series: Effects of Inter-Annual Variability and
Methodology." Remote Sensing 5 (5): 2113-2144.
doi:10.3390/rs5052113.https://www.narcis.nl/publication/RecordID/oai:library.wur.nl:wur
pubs%2F439125.
FREY, D., K. L. H. LEONG, D. FREDERICKS, and S. RASKOWITZ. 1992. "Clustering
Patterns of Monarch Butterflies (Lepidoptera: Danaidae) at Two California Central Coast
Overwintering Sites." Annals of the Entomological Society of America 85 (2): 148-153.
Fung, T and Ledrew, E. 1987.“The application of principal component analysis to change
detection”. Photogrammetric Engineering and Remote Sensing, 53:1649–1658.
Gong, P.1993. “Change detection using principal component analysis and fuzzy set theory”.
Canadian Journal of Remote Sensing, 19:22–29.
Griffiths, Jessica, and F. R. A. N. C. I. S. Villablanca. 2015."Managing monarch butterfly
overwintering groves: making room among the eucalyptus." Cal. Fish & Game 101: 40-
50.

80
Gustafsson, Karin M., Anurag A. Agrawal, Bruce V. Lewenstein, and Steven A. Wolf. 2015.
"The Monarch Butterfly through Time and Space: The Social Construction of an Icon."
Bioscience 65 (6): 612-622. doi://doi.org/10.1093/biosci/biv045
https://academic.oup.com/bioscience/article/65/6/612/302606.
Hartzler, Robert G. 2010. "Reduction in Common Milkweed ( Asclepias Syriaca) Occurrence in
Iowa Cropland from 1999 to 2009." Crop Protection 29 (12): 1542-1544.
doi:10.1016/j.cropro.2010.07.018.
https://www.sciencedirect.com/science/article/pii/S0261219410002152.
Hartzler, Robert G. and Buhler, Douglas D. 2000. "Occurrence of Common Milkweed (
Asclepias Syriaca) in Cropland and Adjacent Areas." Crop Protection 19 (5): 363-366.
doi:10.1016/S0261-2194(00)00024-7.
https://www.sciencedirect.com/science/article/pii/S0261219400000247.

J. B. Wyrill, III and O. C. Burnside. 1977. "Glyphosate Toxicity to Common Milkweed and
Hemp Dogbane as Influenced by Surfactants." Weed Science 25 (3): 275-287.
doi:10.1017/S0043174500033464. https://www.jstor.org/stable/4042790.

Jepsen, S., D. F. Schweitzer, B. Young, N. Sears, M. Ormes, and S. H. Black. 2015.
Conservation Status and Ecology of Monarchs in the United States. 36 pp. NatureServe,
Arlington, Virginia, and the Xerces Society for Invertebrate Conservation, Portl, and Oregon.
http://www.xerces.org/wp-content/uploads/2015/03/NatureServe-Xerces_monarchs_USFS-
final.pdf

John R. Sauer, William A. Link, William L. Kendall, and David D. Dolton. 2010. "Comparative
Analysis of Mourning Dove Population Change in North America." Journal of Wildlife
Management 74 (5): 1059-1069. doi:10.2193/2008-
459. http://www.bioone.org.libproxy1.usc.edu/doi/full/10.2193/2008-459.
Lane, J. 1993. Overwintering monarch butterflies in California: past and present. In Biology and
Con-servation of the Monarch Butterfly, edited by S. B. Malcolm and M. P. Zalucki, 335–
344. Los Angeles: Natural History Museum of Los Angeles County

Leong, K. L. H. 1990. Microenvironmental factors associated with the winter habitat of the
monarch butterfly (Lepidoptera: Danaidae) in central California. Annals of the
Entomological Society of America83:907–910

81
Leong, K. L. H., W. H. Sakai, W. Bremer, D. Feuerstein, and G. Yoshimura. 2004. Analysis of
the pattern of distribution and abundance of monarch overwintering sites along the California
coastline. In The Monarch Butterfly: Biology andConservation, edited by K. S. Oberhauser
and M. J. Solensky, 177–185. Ithaca: Cornell University Press.

Ljiljana, RADIVOJEVIC, SARIC-KRSMANOVIC Marija, GAJIC UMILJENDIC Jelena,
BOZIC Dragana, and SANTRIC Ljiljana. 2016. "The Impacts of Temperature, Soil Type and
Soil Herbicides on Seed Germination and Early Establishment of Common Milkweed."
Notulae Botanicae Horti Agrobotanici Cluj-Napoca 44 (1): 291.
doi:10.15835/nbha44110265. https://search.proquest.com/docview/1801269521.

Liechti, Felix. 2006. "Birds: Blowin’by the Wind?" Journal of Ornithology 147 (2): 202-211.
doi://doi.org/10.1007/s10336-006-0061-9. https://link-springer-
com.libproxy2.usc.edu/article/10.1007/s10336-006-0061-9.

Lo, C. P., and Robert L. Shipman. 1990. "A GIS approach to land-use change dynamics
detection." PE&RS, Photogrammetric Engineering & Remote Sensing 56, no. 1. 1483-1491.

Lu, D., P. Mausel, E. Brondízio, and E. Moran. 2004. "Change Detection
Techniques." International Journal of Remote Sensing 25 (12): 2365-2401.
doi:10.1080/0143116031000139863. http://www-tandfonline-
com.libproxy1.usc.edu/doi/abs/10.1080/0143116031000139863.

Nelson, R. F. 1983. Detecting forest canopy change due to insect activity using Landsat
MSS. Photogrammetric Engineering and Remote Sensing, 49: 1303–1314

Oberhauser, Karen S., Michelle D. Prysby, Heather R. Mattila, Diane E. Stanley-Horn, Mark K.
Sears, Galen Dively, Eric Olson, John M. Pleasants, Wai-Ki F. Lam, and Richard L.
Hellmich. 2001. "Temporal and Spatial Overlap between Monarch Larvae and Corn Pollen."
Proceedings of the National Academy of Sciences 98 (21): 11913-11918.
doi://doi.org/10.1073/pnas.211234298.
http://www.pnas.org/content/pnas/98/21/11913.full.pdf.

Oberhauser, K. and A. T. Peterson. 2003. Modeling current and future potential wintering
distributions of eastern North American monarch butterflies. Proc. Natl. Acad. Sci.
100:14063-14068.

Oberhauser, Karen Suzanne. 2004. Overview of Monarch Breeding Biology
. Monarch Butterfly Biology and Conservation., edited by Karen Suzanne Oberhauser,
Michelle J. Solensky. I ed. Vol. I. New York: Cornell university press.

82

Pearson, Richard G. and Terence P. Dawson. 2003. "Predicting the Impacts of Climate Change
on the Distribution of Species: Are Bioclimate Envelope Models Useful?" Global Ecology
and Biogeography 12 (5): 361-371. doi://doi.org/10.1046/j.1466-822X.2003.00042.x.
https://onlinelibrary.wiley.com/doi/abs/10.1046/j.1466-822X.2003.00042.x

Pelton, Emma and Jepsen, Sarina. "Monarch Butterflies in Western North America in
Jeopardy.", last modified 01/17/, https://xerces.org/2019/01/17/monarch-butterflies-in-
western-north-america-in-jeopardy/.

Pleasants, John M. And Karen S. Oberhauser. 2013. " Insect Conservation and Diversity 6 (2):
135-144. doi:10.1111/j.1752-4598.2012.00196.x.
https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1752-4598.2012.00196.x.p]

Pleasants, John M. And Karen S. Oberhauser. 2013. "Milkweed Loss in Agricultural Fields
because of Herbicide use: Effect on the Monarch Butterfly Population." Insect Conservation
and Diversity 6 (2): 135-144. doi:10.1111/j.1752-4598.2012.00196.x.
https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1752-4598.2012.00196.x.

Pocius, V. M., D. M. Debinski, J. M. Pleasants, K. G. Bidne, R. L. Hellmich, and L. P. Brower.
2017. "Milkweed Matters: Monarch Butterfly (Lepidoptera: Nymphalidae) Survival and
Development on Nine Midwestern Milkweed Species." Environmental Entomology 46 (5):
1098-1105. doi:10.1093/ee/nvx137. https://www.ncbi.nlm.nih.gov/pubmed/28961914.

Rudolph, D. Craig, Charles A. Ely, Richard R. Schaefer, J. Howard Williamson, and Ronald E.
Thill.2006. "Monarch (Danaus plexippus L. Nymphalidae) migration, nectar resources and
fire regimes in the Ouachita Mountains of Arkansas." Journal of the Lepidopterists' Society.
60 (3): 165-170. https://www.fs.usda.gov/treesearch/pubs/all/24947

SINGH, ASHBINDU. 1989. "Review Article Digital Change Detection Techniques using
Remotely-Sensed Data." International Journal of Remote Sensing 10 (6): 989-1003.
doi:10.1080/01431168908903939.
http://www.tandfonline.com/doi/abs/10.1080/01431168908903939.

Stenoien, Carl, Kelly R. Nail, Jacinta M. Zalucki, Hazel Parry, Karen S. Oberhauser, and Myron
P. Zalucki. 2018. "Monarchs in Decline: A Collateral Landscape‐level Effect of Modern
Agriculture." Insect Science 25 (4): 528-541. doi://doi.org/10.1111/1744-7917.12404.
https://onlinelibrary.wiley.com/doi/epdf/10.1111/1744-7917.12404.

83
Stevens, SHAWNA and DENNIS Frey. 2004. "How the Other Half Lives: Monarch Population
Trends West of the Great Divide." Biological Sciences Department, California Polytechnic
State University.Unpublished Report. https://www.monarchwatch.org/update/2004/frey.pdf.

Stevens, Shawna R. and Dennis F. Frey. 2010. "Host Plant Pattern and Variation in Climate
Predict the Location of Natal Grounds for Migratory Monarch Butterflies in Western North
America." Journal of Insect Conservation 14 (6): 731-744. doi://doi.org/10.1007/s10841-010-
9303-5. https://link-springer-com.libproxy2.usc.edu/article/10.1007/s10841-010-9303-5.

Thogmartin, Wayne E., Ruscena Wiederholt, Karen Oberhauser, Ryan G. Drum, Jay E.
Diffendorfer, Sonia Altizer, Orley R. Taylor, John Pleasants, Darius Semmens, and Brice
Semmens. 2017. "Monarch Butterfly Population Decline in North America: Identifying the
Threatening Processes." Royal Society Open Science 4 (9): 170760.
doi:10.1098/rsos.170760. http://rsos.royalsocietypublishing.org/content/4/9/170760.

Urquhart, F. A. and N. R. Urquhart. 1976. "The Overwintering Site of the Eastern Population of
the Monarch Butterfly (Danaus P. Plexippus; Danaidae) in Southern Mexico." Journal of the
Lepidopterists’ Society 30 (AGRIS since : 2012): 153-158.
http://images.peabody.yale.edu/lepsoc/jls/1970s/1976/1976-30(3)153-Urquhart.pdf.

Vidal, Omar, JOSÉ LÓPEZ‐GARCÍA, and EDUARDO RENDÓN‐SALINAS. 2014. "Trends in
Deforestation and Forest Degradation After a Decade of Monitoring in the Monarch
Butterfly Biosphere Reserve in Mexico." Conservation Biology 28 (1): 177-186.
doi://doi.org/10.1111/cobi.12138. https://onlinelibrary-wiley-
com.libproxy2.usc.edu/doi/10.1111/cobi.12138.

Visser, Marcel E. and Christiaan Both. 2005. "Shifts in Phenology due to Global Climate
Change: The Need for a Yardstick." Proceedings of the Royal Society of London B:
Biological Sciences 272 (1581): 2561-2569. doi:10.1098/rspb.2005.3356.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1559974/.

Wassenaar, Leonard I. and Keith A. Hobson. 1998. "Natal Origins of Migratory Monarch
Butterflies at Wintering Colonies in Mexico: New Isotopic Evidence." Proceedings of the
National Academy of Sciences of the United States of America 95 (26): 15436-15439.
doi:10.1073/pnas.95.26.15436. https://www.jstor.org/stable/46395.

Weng, Q. 2002. "Land use change analysis in the Zhujiang Delta of China using satellite remote
sensing, GIS and stochastic modeling ". Journal of Environmental Management, 64: 273–
284.

84
Zalucki, Myron P. 1981. "The Effects of Age and Weather on Egg Laying in Danaus Plexippus
L.(Lepidoptera: Danaidae)." Researches on Population Ecology 23 (2): 318.
doi://doi.org/10.1007/BF02515634. https://link-springer-
com.libproxy2.usc.edu/article/10.1007/BF02515634.

Zalucki, Myron P. 1982. "Temperature and Rate of Development in Danaus Plexippus L. and D.
Chrysippus L.(Lepidoptera: Nymphalidae)." Australian Journal of Entomology 21 (4): 241-
246. doi://doi.org/10.1111/j.1440-6055.1982.tb01803.x.
https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1440-6055.1982.tb01803.x.

Zalucki, Myron P. and Stephen B. Malcolm. 1999. "Plant Latex and First-Instar Monarch Larval
Growth and Survival on Three North American Milkweed Species." Journal of Chemical
Ecology 25 (8): 1827-1842. doi:1020929732223. https://link-springer-
com.libproxy2.usc.edu/article/10.1023/A%3A1020929732223#citeas.

Zipkin, Elise F., Leslie Ries, Rick Reeves, James Regetz, and Karen S. Oberhauser. 2012.
"Tracking Climate Impacts on the Migratory Monarch Butterfly." Global Change Biology 18
(10): 3039-3049. doi:10.1111/j.1365-2486.2012.02751.x.
https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1365-2486.2012.02751.x

85
Appendix A: Monarch Sites and Boundary

Appendix 1: Monarch Sites and Boundary size (example)

86
OBJECTID SITE NAME
1 Andrew Molera State Park
2 Big Creek Preserve
3 Butterfly Grove Sanctuary, Pacific Grove
4 Cabrillo Highway
5 CH1 Private Site
6 CH2 Private Site
7 Deer Flat Ranch
8 Del Monte Road
9 Los Padres National Forest/Caltrans Grove
10 McWay Canyon
11 Moss Landing Middle School, Moss Landing
12 Pacific Grove Golf Course
13 Palo Colorado, South of Carmel Highlands
14 Plaskett Creek, South of Big Sur
15 Point Lobos State Park, Carmel Bay
16 Prewitt Creek
17 Private Property near Big Sur
18 Sycamore Canyon, Big Sur
19 Washington Park, Pacific Grove
20 Yankee Point

Appendix Table 1: Monarch Overwinter Site Locations

Abstract (if available)

Abstract The Western Monarch butterflies of North America have seen a steep decline in population over the last two decades due to climate changes and land cover changes in the overwintering sites, and the decline of milkweed plants in the breeding grounds due to pesticide glyphosate application. According to a study by the Center for Biological Diversity, over the last two decades, monarch populations overall have declined by 68 percent and the western monarchs have seen nearly 90 percent decline, reducing their overall population from more than one billion to two hundred odd million (Giffiths & Villablanca, 2015). ❧ This project aimed to quantify the land use land cover changes, trend of pesticide Glyphosate application from 1990 to 2015, weather changes from 1990-2015 in the Monterey county, the study area for this project. The land cover and land use change were analyzed using Landsat ARD imagery of 1994, 1998, 2002, 2006, 2014 and 2018, using NDVI (Normalized Difference Vegetation Index) differencing method. The NDVI image for each year were classified into 5 categories based on the spectral index and the results indicate fluctuation in the land use land cover, however, there is not significant change in any class. The pesticide Use Report showed Grapes and uncultivated agricultural land used high amount of pesticide glyphosate each year. The application had increased from 1996 to 2000 and again it is increasing from 2004. The weather data analysis did not show any inclement weather pattern, which might be unsuitable for monarch winter survival.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

A comparative study of ground and satellite evapotranspiration models for southern California

PDF

A comparison of urban land cover change: a study of Pasadena and Inglewood, California, 1992‐2011

PDF

Integrating Landsat and California pesticide exposure estimation at aggregated analysis scales: accuracy assessment of rurality

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

Assessing woody plant encroachment in Marin County, California, 1952-2018

PDF

Using Landsat and a Bayesian hard classifier to study forest change in the Salmon Creek Watershed area from 1972–2013

PDF

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

PDF

Using Maxent modeling to predict habitat of mountain pine beetle in response to climate change

PDF

Developing a replicable approach for the creation of urban climatic maps for urban heat island analysis: a case study for the city of Los Angeles, California

PDF

Classifying the alpines: developing a methodology to track environmental changes in the alpines utilizing remote sensing and ecozone vegetation patterns

PDF

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

PDF

Historical ecology of the Split Oak Forest in east central Florida

PDF

Using census data, urban-land cover classification, and dasymetric mapping to measure urban growth of the lower Rio Grande Valley, Texas

PDF

Mapping future population impacts caused by sea level rise in Huntington Beach and Newport Beach: comparing the cadastral-based dasymetric system to past dasymetric mapping methods

PDF

Spatiotemporal analysis of the SLOSH and ADCIRC storm surge models: a case study of hurricane Ida

PDF

Modeling burn probability: a Maxent approach to estimating California's wildfire potential

PDF

Identifying suitable sites for sheltering outside in Long Beach, California

PDF

Effects of building modifications and municipal policies on green cover in Los Angeles County

PDF

Investigating bus route walkability: comparative case study in Orange County, California

PDF

Assessment of land cover change in Southern California from 2003 to 2011 using Landsat Thematic Mapper

Asset Metadata

Creator Hulisandra, Indu R. (author)

Core Title Quantifying changes in monarch butterfly habitat in California

School College of Letters, Arts and Sciences

Degree Master of Science

Degree Program Geographic Information Science and Technology

Publication Date 03/23/2020

Defense Date 12/17/2019

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

Tag California,herbicide glyphosate,land use land cover changes,Landsat ARD,Monarch butterfly,monarch butterfly yearly count,Monterey County,North American monarch butterfly,OAI-PMH Harvest,overwintering sites,spatio-temporal change analysis,weather changes,western monarch butterfly

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Bernstein, Jennifer (committee chair), Marx, Andrew (committee member), Ruddell, Darren (committee member)

Creator Email induhulisandra@yahoo.com,indujoshi@icloud.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-276781

Unique identifier UC11673092

Identifier etd-Hulisandra-8228.pdf (filename),usctheses-c89-276781 (legacy record id)

Legacy Identifier etd-Hulisandra-8228.pdf

Dmrecord 276781

Document Type Thesis

Rights Hulisandra, Indu R.

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