TABLE OF CONTENTS
DEDICATION i
ACKNOLEDGEMENTS ii
ABSTRACT iii
LIST OF FIGURES v
LIST OF TABLES vii
LIST OF EQUATIONS vii
INTRODUCTION AND BACKGROUND
Fate of Pharmaceuticals 1
Wastewater Treatment Facilities 2
Alternate Uses for WWTP End Products 4
Carbamazepine 6
Primary Producers- Tomatoes 8
Primary Consumers- Hornworms and Moths 10
Better Technology for Environmental Pharmaceutical Detection 13
Significance 14
Hypothesis and Specific Aims 16
EXPERIMENTAL METHODS
Experimental Design 16
Chemicals 17
Treatment of Samples 18
Plant Growth 19
Rearing Worms and Moths 21
Sample Analysis 22 Instrumental Analysis 26 Cleaning Protocols 29 Quality Control 30 Statistical Analysis 31 RESULTS Statistical Analysis 32
Tomato Plant Data 33
Hornworm Data 35
Hawkmoth Data 37
CBZ Transfer 38
Worm Excrement Data 41
DISCUSSION AND CONCLUSIONS
Tomato Plants 42
Hornworms 42
Hawkmoths 43
CBZ Transfer 44
Worm Excrement 46
LITERATURE CITED 48
i
DEDICATION
No bountiful journey is ever walked alone and no fruitful labors are ever achieved without the
assistance of kind souls and so I would like to dedicate my thesis dissertation to my dearest
family for without whom none of this would be possible. They are my surest road and a lasting
joy along the journey of my life.
&
ii
ACKNOWLEDGEMENTS
I would like to extend my deepest gratitude to everyone who has helped me along the way.
THANK YOU! Dr. Brian, thank you for all of your help and assistance as my graduate advisor
and setting me on a project that gave me many more experiences that I thought I would have
never had. Dr. Kinney, thank you for your immense patience with me and for helping me
through all things chemistry. I am very grateful for all of your time and efforts. Dr. Claire, Thank
you for your great statistics help and for walking me through each stats aspect. To my dearest
family, thank you for your love and support the entire time I was working on my project. Dad,
you built the best hornworm hutches and helped me so much in keeping the tomato plants,
hornworm and hawkmoths alive. I am forever grateful! Mom, thank you for your wonderful help
when I needed it the most to work through the problems and situations that came up during my
project. Your great love and tremendous care are what got me through. Lauren, I would have
been lost without you. You helped me figure out and complete so many chemistry tasks! I could
not have done it without you; thank you. Nicole, thank you for giving me encouragement and
reminding me to keep working even when I felt like quitting. Ryan, thank you for everything you
have done for me. You have been a great friend and wonderful support to fall back on. I am so
grateful that you were always there for me and I will never forget that. You are the best lab/
project partner I could have ever had. Theresa, thank you all for all of your dedication to me
these past five years as I have grown and progressed and am now finished at the CSU-Pueblo
Life Science Department. I am truly grateful for all you have done for me! By the grace of God
iii
Abstract
Bioaccumulation of Carbamazepine in Hornworms by Herbivory of Tomato Plants Grown in Soil Spiked with Carbamazepine
There is an increasing trend in agriculture to use biosolids as a means of replenishing nutrients to
crop fields. Biosolids are treated sewage sludge obtained from waste water treatment plants and
have been reported to contain active pharmaceutical compounds that survive the wastewater
treatment process. When the biosolids are used to amend fields, the active pharmaceuticals are
also transferred to the agricultural fields where they may leach though the soil system into
ground water, may accumulate in terrestrial organisms such as the crop plants growing in the
amended fields or undergo biotic and abiotic transformation processes. A growing body of
research has focused on pharmaceuticals being transferred to crop plants; however there is not as
much describing whether the pharmaceuticals are further transferred through terrestrial
ecosystems to higher trophic levels. Carbamazepine (CBZ) is a commonly found pharmaceutical
in biosolids and has been shown to be taken up by a variety of crops including tomatoes. This
study investigates the transfer of CBZ from tomato plant leaves to hornworms to illustrate the
potential for pharmaceuticals to be transferred through increasing trophic levels. Tomatoes
grown under greenhouse conditions in CBZ spiked soil were used as food for hornworms that
were collected for analysis at the fifth instar and also collected after pupation as adult
hawkmoths. In this study CBZ was detected in the tomato leaves (184.77 ng/g dw), hornworm
tissue (27.48 ng/g dw) and hawkmoth tissue (3.76 ng/g dw). CBZ present in the hornworm and
hawkmoth tissues demonstrates the potential for transfer of pharmaceuticals from plants through
iv
remained present in them through metamorphosis meaning the compound was sequestered and
remained in their tissues. Although not investigated as part of this study, this may lead to
negative behavioral effects. The decreasing concentration trend from plant to hornworm to
hawkmoth suggests that bioaccumulation and biomagnification are not occurring within these
organisms and likely suggests metabolism/ elimination in the organism. Further inquiries are
needed on movement and fate of pharmaceuticals through trophic levels; for example,
metabolites of CBZ could be measured in hornworm and hawkmoth tissue to see if the
compound is being metabolized. This research could help to enlighten the fate of CBZ and other
v
List of Figures
Figure 1 3
Wastewater treatment process with primary and secondary treatment. Photo courtesy of lenntech.com/wwtp/wwtp-overview.htm
Figure 2 6
Chemical Structure of Carbamazepine; SigmaAldrich.com
Figure 3 7
Various CYP pathways for the breakdown of carbamazepine into its metabolites and some end products (Pearce et. al., 2008). Circled metabolites are discussed
Figure 4 11
Hornworm life cycle. Hornworms begin life as a free egg and grow through 5 instars before they reach a pupa stage and finally the adult moth stage. Photo courtesy of rainbowmealworms.net
Figure 5 17
Flow chart of experimental design. The Carrier solvent is methanol (MeOH) and the treatment compound is carbamazepine (CBZ). The negative control group was not exposed to carrier solvent or CBZ. No CBZ is expected to be detected in this group. The carrier control group was exposed to the carrier solvent and no CBZ is expected to be detected with this group. The CBZ treatment group that was exposed to both carrier solvent and CBZ is expected to be detected in the samples.
Figure 6 20
This picture shows the hutches that were built to house individual tomato plants, hornworms and moths.
Figure 7 24
This picture shows the ASE instrument that was used to complete all of the sample extractions in this study.
Figure 8 26
This picture shows the UHPLC MS/MS that was used to analyze all of the samples qualitatively and quantitatively for carbamazepine.
Figure 9 33
This graph displays back transformed concentrations for plant, hornworm and hawkmoth
samples for control and CBZ treatment groups. Negative and carrier control group samples were combined for statistical analysis. (*) indicates statistical differences between treatment groups. (**) indicates statistical differences between organism groups. Two-way ANOVA gave a p-value of 2.20x10-16 for the control and CBZ treatment groups and a p-value of 4.21x10-4 for differences between organisms at 95% confidence level (α= 0.05).
vi
Figure 10 35
The average CBZ concentrations (on a dry weight basis) of the plant samples for all three
treatment groups and has not been manipulated for statistical analysis. NC is the negative control group. CC is the carrier control group and T is the CBZ treatment group. Error bars represent standard error of the mean.
Figure 11 36
The average CBZ concentrations in the hornworm samples for all three treatment groups and has not been altered for statistical analysis. NC is the negative control group, CC is the carrier
control group and T is the CBZ treatment group. Error bars represent standard error of the mean.
Figure 12 38
The average CBZ concentrations of the moth samples for two treatment groups and have not undergone transformation for statistical analysis. CC is the carrier control group and T is the CBZ treatment group. There were no moth samples for the negative control group. Error bars represent standard error of the mean.
Figure 13 39
The translocation factor (TF) for the plant, worm and moth samples. TF (P:W) is the transfer of CBZ from plant to hornworm and TF(P:M) is the transfer of CBZ from plant to hawkmoth.
Figure 14 40
The concentration of CBZ in each sample of the three different organisms. This data is from the CBZ treatment group only. Sample numbers 1, 4, 5, 8 and 9 display the decreasing CBZ
concentration trends from all three organisms. The graph shows transfer of CBZ from plant to worm to moth.
Figure 15 40
Scatter plot of the tomato plant and hornworm samples CBZ concentration showing a positive correlation with regards to CBZ concentrations in plant and worm samples.
Figure 16 41
This graph shows the overall concentration of CBZ in worm excrements. NC is the negative control, CC is the carrier control and T is the CBZ treatment group.
vii List of Tables
Table 1: Selected Physical and Chemical Properties of Carbamazepine 6
Table 2: Parameters of tandem Mass Spectrometer determined during method development for
detection and quantification of carbamazepine. 27
Table 3: Plant Samples for all treatment types and individual sample numbers (ng/g) 34
Table 4: Worm samples for all treatment types and individual sample numbers (ng/g) 36
Table 5: Moth samples for all treatment types and individual sample numbers (ng/g) 37
List of Equations
1 Introduction and Background
Fate of Pharmaceuticals
There is increasing use of pharmaceuticals in the United States and worldwide (Richards
2010). America is one of the largest consumers of pharmaceuticals which have become a
necessity in life to improve health and better patient outcomes (O’Neill & Sussex 2014; WHO 2004). Pharmaceuticals are metabolized in humans mainly by liver enzymes. However,
incomplete metabolism results in the excretion of active pharmaceuticals and subsequent release
into waste water treatment systems. Other modes by which active pharmaceuticals can enter
waste water systems include disposal of leftover or expired pharmaceuticals down the drain or
flushed down the toilet, as well as accidental release into the environment following
pharmaceutical production (USGS 2009). Pharmaceuticals in public water systems eventually
can be introduced into the natural environment through land application of Waste Water
Treatment Plant end products.
Legislation and regulations exist for most persistent organic contaminants like pesticides
and industrial by-products (EPA 2009; Muir & Howard 2006), but there is minimal regulation
for pharmaceutical pollution. Before a pharmaceutical is allowed to be marketed it is required to
undergo Environmental Risk Assessment which determines the concentration at which it could
potentially cause harm to the natural environment (Meisel et al. 2009). The pharmaceutical
compounds entering the environment are commonly known as emerging pollutants or
micro-pollutants (Garcia-Rodriguez et al. 2014; Farré et al. 2008; Vodyanitskii & Yakovlev 2016).
Currently, the EPA is assessing pharmaceutical threat in the environment but no definite actions
2
environment (Cotruvo 2012). Pharmaceuticals are considered “chemicals in commerce” but have
not yet been included and considered part of the persistent organic pollutants which are synthetic
organic chemicals such as pesticides and industrial chemicals as defined by the World Health
Organization (Muir & Howard 2006; Damstra-WHO 2008). Pharmaceuticals however meet the
criteria as described by Muir and Howard (2006), which are as follows: they are well
characterized for toxicity, are entering the environment and are not being metabolized or
degraded quickly.Pharmaceutical compounds being introduced into the environment is a
relatively new branch of research that has the potential to help legislation and the public gain a
better understanding of the environment surrounding them. With a better understanding of how
pharmaceuticals interact in the natural environment, legislators can make better regulations that
will further protect public health and safety.
Waste water Treatment Facilities
Good health and clean cities require the removal of human waste products (Henze et al.
2008). As the world population began to grow exponentially at the turn of the twentieth century,
more waste products were being produced. Public officials and civil engineers were faced with
the challenge of removing the wastes to provide their citizens with clean and safe places to live.
Towards the end of the nineteenth century the first wastewater treatment plants were
implemented. These “treatment plants” were operated by having citizens collect and take their
wastes to cesspools or privies (Burian et al. 2000). This became an irritation for people and as
time went on more modern facilities were designed to remove the sewage that often caused
disease, and released usable water back into the environment. Wastewater treatment plants are an
3
to WWTP via the underground sewer system. Sewage sludge and treated waste water effluent are
the end products of this process (Figure 1). During waste water treatment large particulates are
filtered out either by skimming or settling. Then microbes are allowed to breakdown suspended
particulates and decrease the biological oxygen demand. These first two components are the
primary and secondary treatments required by law in order to put the sewage water back into
natural water systems (EPA 1998). Some WWTP also have tertiary and quaternary treatments
such as chlorination to further disinfect the sludge and effluents but they are expensive and so are
not always implemented. At the end of the waste water treatment process, there is leftover
organic matter called sewage sludge. The sludge can become biosolids if it meets standard
regulations for disinfection of pathogens and heavy metals and biosolids can be utilized in a
variety of ways (Stevens 2016). Sewage sludge is commonly disposed of by incineration,
dumped in landfills or used on agriculture soil. The left over components must be disposed of
properly and efficiently so that treatment plants can continue to run smoothly and remain within
regulatory guidelines. Other treatment processes include biologically active waste water
treatment. Current studies are conducting research to effectively remove pharmaceuticals like
CBZ from effluents of WWTP. One study specifically is using the electro-Fenton process in
4
retrieves electrons from pharmaceuticals like Carbamazepine (Komtchou et al. 2015). Another
study is looking at the use of Microalgae as a tertiary treatment that can help to remove the
excess of nitrogen and phosphorus compounds in wastewater effluents. They also have the
potential to take up both organic and other inorganic pollutants (Abdel-Raouf et al. 2012).
Alternate Uses for WWTP End Products
A more useful means of disposing it is application to nutrient deficient or nutrient
depleted soils because it is a rich source of carbon and nitrogen, potassium and phosphorus
which are important elements for plant growth. Approximately 250 metric tons of biosolids are
added to soil in the United States every year (McClellan & Halden 2010). This can include crop
fields and reclamation sites like deforested areas and mine tailing dump sites. There are benefits
to using biosolids as fertilizer for these sites because of their lower cost, lower potential for
contamination of water ways, better availability for plant uptake of nutrients, enhancement of top
soil and an increase of soil organic matter (McIvor et al. 2012; Hughes & MacKay 2011). Many
pharmaceuticals including antibiotics, anti-convulsants and many other types have been found in
agricultural soils (Kinney et al. 2008; Al-Rajab et al. 2015; EPA 2010; EPA 2011). WWTP are
not able to remove all of the pharmaceutical compounds from the end products so these stable
compounds are released into aquatic environments in treated effluent and terrestrial
environments through land application of biosolids and irrigation with reclaimed waste-water
(Uslu et al. 2013). Carbamazepine is one of the pharmaceuticals that is not completely removed
during the treatment process. Because pharmaceuticals remain in the biosolids they are also then
introduced to the environment by land application and are considered emerging contaminants
5
their tissues. The physiochemical properties of pharmaceuticals play a role in how much and
what types of organisms take them up (Hughes & MacKay 2011). Pharmaceuticals can be
metabolized and metabolites can be potentially harmful (Farré et al. 2008). Other than in
agricultural application biosolids are used for amending soil in forests, range lands and
reclamation sites (EPA 2010; EPA 2011). Mine tailing sites have long been a focus for needed
reclamation. Biosolids have long lasting effects at these sites even when they are applied just one
time (Pepper et al. 2012) so it is important to know the safe level for land application of biosolids
in an area. Biosolids have been reported to lower soil pH (McIvor et al. 2012) and soil chemistry
is important for good plant growth and should be considered when applying biosolids to any site
in order to not damage an area from application of biosolids (Paschke et al. 2005). Some research
has suggested that there are potential health risks because of biosolids amendment like microbial
infection (Brooks et al. 2012; Lowman et al. 2013; Robinson et al. 2012). Land application of
biosolids (when used with caution) is an economic way of disposing of biosolids but more
information must be obtained to understand the fate of pharmaceuticals now being introduced
into the environment and their effects on other organisms.
The liquid effluent that comes out of the wastewater treatment plants is an important
source of reclaimed water. If it is not released back into the environment it can be used to irrigate
agricultural fields in arid and semiarid zones (Shenker et al. 2011). It can also replenish depleted
water resources. Use of waste water is important in places where water resources are scarce.
Pharmaceuticals have been shown to remain in the effluent water providing another mode by
which these compounds are introduced into the natural environment either through crop
6
negative effects on crop production and public health and safety according to some research
(Christou et al. 2014). However it is still important to understand where pharmaceuticals and
other emerging contaminants end up and whether they arecausing harm or not within the natural
environment.
Carbamazepine
A commonly detected pharmaceutical in Wastewater Treatment Plant (WWTP) effluents
is Carbamazepine (CBZ; Mohapatra et al. 2014). It is a tricyclic compound (Table 1; Figure 2)
used as an anti-convulsive and for trigeminal neuralgia (Bertilsson 1978). Its brand name is
Tegretol but there are generic brands of the drug available on the pharmaceutical market. Even
though these generic forms of the drug are commonly used there is evidence to support that
people encountered negative effects when switching from the brand product to the generic forms
(Tothfalusi et al. 2008). Carbamazepine is also used as a treatment for Bipolar Disorder and
depression (Paz et al. 2016). It has been on the market for many decades now and is commonly
prescribed nationally and internationally. The LD50 oral dose in rat is 1957 mg/kg and it has
been shown to cause reproductive toxicity (SDS Sigma-Aldrich 2016). The main mechanism of
7
human body by the liver enzyme Cytochrome P450 (CYP3A4; Myllynen et al. 1998).
Carbamazepine was used in this study because of its consistent presence in WWTP end products,
such as reclaimed waste water and biosolids (Bosch et al. 2014) and this makes it an ideal model
pharmaceutical to study with regards to environmental fate.
There are many metabolites of CBZ, some of which have known toxicity (Figure 3).
Carbamazepine2,3 epoxide is potentially harmful because it can covalently react with
macromolecules like proteins and lipids and would stop normal functions of cells (CMC 2013).
Carbamazepine 10,11- epoxide does not seem to possess any toxic effects in humans and still has
psychoactive properties (Valentine et al. 1996; Pirmohamed et al. 1992). 10,11-dihydro-trans-10
11-dihydroxy carbamazepine is another metabolite of the parent compound and does not have
any known toxicity nor does it have psycho-active properties. Studies have been conducted that
8
(2015) discussed negative effects on bivalve clams when they were exposed to carbamazepine
chronically. The extended exposure weakened their immune systems which lead to an overall
decrease in the viability of the clams. The study also proposed a trend of concentration
dependent effects for the organisms. As CBZ concentrations increased, mortality increased.
Carbamazepine is a relatively non-polar, neutral compound and often can accumulate in lipid
regions of organisms because of its conjugated aromatic rings (Ungureanu et al. 2015).
Carbamazepine that accumulates in plants has been found to mainly remain in the leaves of
plants and water flow transport is considered the main mechanism for translocation of CBZ
throughout the plants (Shenker et al. 2011). Various reports have shown that CBZ concentration
levels range from the low parts per million (µg/g) to the low parts per billion (ng/g) in aquatic
and terrestrial environments (McClellan and Halden, 2010; Miao et al. 2005; Ungureanu et al.
2015; Uslu et al. 2013).
Primary Producers- Tomatoes
Ecosystems are filled with plant and animal relationships in which herbivores depend on
primary producers to survive These predator-prey (plant-herbivore interactions) relationships
help to increase adaptation and natural change throughout the species of both the plants and
animals as they try to survive better than the other (Labandeira et al. 2016). Certain animals like
insects are specialists meaning they depend on particular plants for food (Bosch et al. 2014).
Plants have various defense mechanisms that these specialists must overcome in order to
continue growth and development over their life cycle. There can be negative effects within
entire ecosystems if these relationships are interrupted either by one species not being available
9
occurring phenomena (Part 503, EPA 2016 A). Tomatoes are a very important crop worldwide.
They are a member of the Solanaceae family of plants and have many predators which threaten
crop production.
Many factors effect crop plant production including biotic and abiotic stresses. Plants in
the natural environment encounter many pests who can be very destructive by consuming the
plants or passing along plant diseases. They are often also exposed to heavy metals and other
organic and inorganic contaminants that can potentially have negative effects on the plants and
therefore crop production. There have been many studies conducted to test the stress of drought,
salinity and pesticides on these plants; however few studies have addressed the effects of other
organic compounds, like pharmaceuticals, on tomato plants. High crop yield is the main goal of
commercial farmers and so it is important to know how pharmaceuticals introduced into
agro-ecosystems by irrigation with reclaimed wastewater or application of biosolids as soil
amendment are affecting the plants and whether they are being taken up by the fruits passed on
to consumers. According to University of Georgia Extension tomatoes are an extensively grown
crop in the United States (Commercial Tomato Production Handbook, UG Extension 2017). The
production of tomatoes requires a very detailed management plan that includes pest control and
fertilization. Tomatoes feed a great number of people nationally and internationally. Tomato
production is also very important economically with largest exports going to Canada
(Commercial Tomato Production Handbook, UGA Extension 2017). Tomato crop production
employs over 10 million people in Mexico alone with tomatoes being their most important
economic export (Medina-Saavedra et al. 2017). Tomatoes are used in the gastronomy of many
10
Some studies have shown that plants (especially crop plants) like carrots, cabbage, spring
wheat and soybeans take up CBZ (Carvalho et al. 2014; Christou et al. 2017; Prosser et al. 2014;
Holling et al. 2012). The compound can be found in roots, stems and shoots depending on the
plant. Tomatoes take up carbamazepine at a wide range of concentrations and with variability
between roots, stems and leaves (Zheng et al. 2014; Paltiel et al. 2016; Sheikh 2017; Chefetz et
al. 2015). CBZ is most often found at the highest concentrations in the leaves; which suggest that
it is being transported by mass water flow through plants (Carvalho et al. 2014).
Primary Consumers- Hornworms and Moths
Hornworms (Manduca sexta (Linnaeus 1763)) make a good indicator species because of
their larger size and ability to be easily reared in laboratory settings. The life cycle of hornworms
can help researchers better understand the effects conservation of pharmaceuticals has on
metamorphosing insects (Diamond et al. 2010). These hornworms are often considered crop
pests and there is great effort put into controlling them. They can cause overwhelming damage to
crops of the Solanaceae family which often is their main source of food. These include tobacco,
potatoes, tomatoes and ornamentals like Angel’s Trumpet and Petunias. Plants in this family can
produce secondary metabolites (mostly sterol derivatives) with insecticidal properties
(Weissenberg et al. 1998). Hornworms feeding on tomatoes are not affected by the secondary
metabolites the tomatoes produce (Ventrella et al. 2016). They have adapted to metabolize these
compounds from plants with varying plasticity of ability (Pauchet et al. 2010). However they are
not necessarily able to metabolize other xenobiotics that are not naturally occurring within the
plant-herbivore relationship. Introducing synthetic chemicals into their food may cause negative
11
the ability to perform secondary metabolism that can help to eliminate xenobiotics. It is not
known though if CBZ is metabolized in hornworm digestive system the same way it is in
humans, the target organism for this compound.
Normal life cycles of insects may be affected by the food they consume if it has toxic
metabolites in present in it (Machdo et al. 2015). Hornworms hatch from eggs and grow through
5 larval instars before they go through metamorphosis to become adult hawkmoths (Figure 4).
The environment that the eggs are laid in greatly affects the hornworm’s ability to prosper
through their larval and adult stages (Potter et al. 2011). The larva emerges from the egg as a
completely functional individual and immediately begins searching for food. This is the first
growth stage, called the first instar, in which the hornworms are approximately 3mm in length.
At each instar they shed their outer cuticle so new larger one can form. Near the fifth instar, the
hornworms increase their food consumption greatly in order to have enough energy stored for
12
source to find a place to burrow underground. They make a burrow structure to help protect them
from unfavorable environmental conditions in their surroundings (Sprague & Woods 2015).
Once the worm is encased in its burrow a chrysalis forms around the changing body of the worm
and complete metamorphosis occurs during a time period of approximately three weeks. The
hornworms have many predators like birds and wasps. Coccinellidae (ladybugs) will feed on
hornworm eggs. Braconid wasps parasitize the hornworms by laying their eggs on the
caterpillars and the larvae feed on the worm when they hatch. Hornworms have adapted in
various ways to protect themselves from predators. Often they will feed less in the presence of
predators to decrease exposure but in doing so their metabolic activity also changes (Thaler et al.
2014). Their morphological and genetic traits help them to continue growing and surviving even
under suboptimal conditions.
As the hornworms complete metamorphosis, they emerge as moths, commonly known as
Hawkmoths, Hummingbird Moths or Sphinx Moths. These moths live for several weeks during
which they feed, breed and reproduce. Because of their feeding habits, they are considered
specific plant pollinators. Once the moths emerge from their underground burrows they begin
forging for flower nectar which is their main source of energy. They have the ability to hover
over their food sources, similarly to hummingbirds and so they track the motion of flowers as the
flowers move with the breeze. This behavior requires them to consume large amounts of food on
a constant basis (Campos et al. 2015). These moths often feed from trumpet-shaped flowers and
flowers with this morphology include genera like Datura and Nicotiana and many other species
in the Solanaceae family. Hawkmoths have an important plant-pollinator relationship with this
13
proboscis is guided in towards the center of the flower due to the curves of the corolla (Campos
et al. 2015). As the moth makes contact with the flower, pollen is able to transfer to the moth and
then to other flowers of the same species as the moth moves to other flowers to obtain nectar.
Flowers of this family include many important crop plants, aesthetic plants and wild plants who
all contribute to a functional ecosystem. Hawk moths also use these plants for reproduction.
Female moths do not have a specific pattern when laying their eggs on leaves of plants (Potter et
al. 2012). Reproductive success depends on the mortality of the offspring; the ability of the
hornworms to hatch from their eggs and then survive in environmental conditions.
Better Technology for Environmental Pharmaceutical Detection
Ultra-High Performance Liquid Chromatography Tandem Mass Spectrometry
(UHPLC/MS/MS) was used in this research to analyze CBZ in plant, moth and worm tissue both
qualitatively and quantitatively. This analysis technique has become a standard for analysis of
trace organic compounds in bioanalytical research as well as an important technology for the
pharmaceutical industry. UHPLC/MS/MS has high sensitivity and selectivity (Naxing Xu et al.
2007), which is very useful in analyzing trace compounds in biological and environmental
system. It can provide better resolution of instrumental response peaks when analyzing multiple
analytes in biological samples which significantly helps diminish matrix effects (Chambers et al.
2007). To gain better resolution a small particle size is used (sub-2µm particles) and allows for
the use of higher pressure which decreases analysis time. This technology can help researchers
better understand environmental concentrations of contaminants in the environment by providing
them with faster analysis time and improved separation of analyses in complex matrices (Mazzeo
14
compounds like pharmaceuticals or other trace organic contaminants in environmental matrices
can be effectively conducted.
Significance
Agricultural land is commonly amended with biosolids as an inexpensive and nutrient-
and organic carbon- rich fertilizer (the mode by which pharmaceuticals can be introduced into
the environment). Tomatoes (or other various crop plants) planted in fields amended with
biosolids are then potentially exposed to pharmaceuticals and other contaminants in the biosolids
(Christou et al. 2017; Sridhar et al. 2014). Often plants, like tomatoes, have the ability to take up
and accumulate the pharmaceuticals in their tissues (Paltiel et al. 2016; Bahman 2017; Zheng et
al. 2014). Other studies have demonstrated that pharmaceuticals can be metabolized by plants
(Carvalho et al. 2014). Pests, like hornworms, who invade crop fields by their natural life
processes, eat the plant material and the hornworms then can bioaccumulate pharmaceuticals in
their tissues. Kinney et al. (2012) has shown pharmaceutical uptake in earthworms from
biosolids amended soils. These compounds are not intended for organisms like hornworms and
therefore when non-target organisms are exposed to pharmaceuticals like CBZ there is potential
for negative effects on exposed individuals. This can potentially change their behavior so that
their natural life cycle is disrupted (Tierney et al. 2014). Hornworms and the moths they become
have an important role in natural ecosystems (even though they are considered crop pests) by
being food for birds and other predators and pollinating specific plant genera. Transfer of the
compound can potentially go even further up the food chain to higher tropic levels. Sherburne et
al. (2016) showed transfer of the antimicrobial drugs Triclosan and Triclocarban, being passed
15
When synthetic substances enter the natural environment there is always concern for the
potential harm they can do. Human pharmaceuticals (not intended for non-human organisms) are
entering the environment and are being taken up by the foods people consume (Paltiel et al.
2016). Aside from human consumption there is little knowledge on the movement of these
compounds to higher tropic levels. Because they are still active, pharmaceuticals may have
negative effects on the organisms that also consume the plants or have a mutualistic relationship
with them (McClellan & Halden 2010).
Bioaccumulation in hornworms could cause additional unintended problems such as
biomagnification in the organisms who prey on the hornworms. This concept is not completely
new however. The threat that DDT caused to Brown Pelicans, Peregrine Falcons and Bald Eagles
became a devastating reality when their species suffered from reproductive failure (Hellou et al.
2013). Some pharmaceuticals do not degrade quickly and so remain within a natural environment
for long periods of time along with continual application of biosolids and thus pharmaceuticals.
By land application of biosolids or irrigation with reclaimed wastewater, pharmaceuticals are
being introduced into the environment at low concentrations. This chronic exposure to various
compounds can potentially have negative effects (Boxall et al. 2012; Fatta-Kassinos et al. 2011;
Vasquez et al. 2014; Zenker et al. 2014; Christou et al. 2017). It is important to understand
where pharmaceuticals are ending up and this study aims to investigate the potential
bioaccumulation of CBZ in hornworms which will help to provide a stepping stone for further
16 Hypothesis and Specific Aims
Hypothesis: Carbamazepine will be present in hornworm tissue after they have eaten
foliage of tomato plants grown in soil fortified with environmentally-relevant concentration of
CBZ and will remain present through pupation from hornworm to hawk moth.
Specific Aim 1a: I aim to determine whether CBZ will be present in hornworm tissue and
the concentration of CBZ in hornworms.
Specific Aim 1b: I aim to determine whether CBZ remains present in moth tissue after
metamorphosis of hornworm and the concentration of CBZ in the moths.
Methods
Experimental Design
Three sets of tomato plants, hornworms and moths were grown. The first set was the
negative control in which plants were given neither treatment compound nor carrier solvent. The
second set was a mixed set of carrier controls and treatment plants. The carrier control plants
contained only solvent and the treatment plants were given both solvent and treatment
compounds. The third set was identical to the second set. Soil was either spiked with treatment
compound and solvent, just solvent or nothing by filling the pots and then mixing (Figure 5).
Methanol was used as the carrier solvent. 0.134 µL of CBZ was aliquoted from a 1 mg/mL
solution into 300 mL of methanol and the solution was poured into a bucket with growing matrix
from the planting pots so that 0.13 mg of CBZ would be in each pot to produce a final
17
by hand and then placed back into the pot. The soil was allowed to equilibrate for 48 hours in the
greenhouse prior to planting. All samples were harvested according to maturity of individual
organism. Tomato plant tissue was collected at approximately 120 days and hornworm and
hawkmoth samples were collected at approximately 30 days. All samples were analyzed on an
Ultra- High Performance Liquid Chromatograph tandem Mass Spectrometer (UHPLC/MS/MS)
instrument. Samples underwent statistical analysis to find significant differences in control and
treatment samples.
Chemicals
Carbamazepine was obtained from Sigma-Aldrich Inc. Isotope-labeled carbamazepine-
13
C-6 was obtained from Cerilliant Sigma-Aldrich Co and came in a one-time use ampule (100
µg/mL in 1 mL methanol). Standard Solutions of both the unlabeled and isotope-labeled Figure 5. Flow chart of experimental design. The Carrier solvent is methanol (MeOH) and the treatment compound is carbamazepine (CBZ). The negative control group was not exposed to carrier solvent or CBZ. No CBZ is expected to be detected in this group. The carrier control group was exposed to the carrier solvent and no CBZ is expected to be detected with this group. The CBZ treatment group was exposed to both carrier solvent and CBZ is expected to be detected in the samples.
18
compounds were prepared for use using a 50/50 water methanol solvent mixture. All solvents
used including acetonitrile, methanol and water were Optima grade (for UHPLC use) except for
the dichloromethane as part of the extraction solvent mixture which was HPLC grade. Chemical
stock solutions were stored at -20 ⁰C and standards and sample solutions were stored in the refrigerator. Alumina (Aluminum Oxide) was obtained from ‘Baker Analyzed’® Reagent and
was a neutral compound that was Grade I, for Chromatography. Silica gel was obtained from
Alfa Aesar and was 70 to 200 mesh powder. Ottawa sand (SiO2) was obtained from BDH®
VWR® Analytical.
Treatment of Samples
The amount of CBZ used to spike the soil matrix was based on reported CBZ
concentrations in biosolids (Miao et al. 2005) and biosolids application rates. Miao et al. (2005)
analyzed biosolids and wastewater from the Peterborough Ontario, Canada WWTP to detect
carbamazepine and five of its metabolites. They found a rage of CBZ concentrations in the
biosolids. In a study by McClellan and Halden, (2010) CBZ concentration was reported for a
wide variety of pharmaceuticals, personal care products and other organic contaminants in
biosolids. These researchers analyzed 110 composite biosolids samples that were taken from 94
Waste Water Treatment Plants in 32 States as part of the 2001 EPA National Sewage Sludge
Survey. They reported an average concentration of CBZ to be 163 mg/kg. They also reported
that CBZ was detected in all of the samples tested. For this study the Land Application Guide
chart provided by the EPA, (EPA, 2000), at an agricultural application rate of 20 dry
tons/acre/year was selected to base the soil spike on. This rate was brought to scale for the
19
planting pot. This lower concentration is environmentally relevant because it is within a range of
other reported CBZ concentrations in biosolids (Miao et al. 2005; McClellan & Halden 2010). It
also ensured organism survival in that CBZ can be more toxic then other pharmaceuticals during
exposure of non-target organisms and can have negative impacts on plant development (Farré et
al. 2008; Carter et al. 2015).
Plant Growth
Tomato plants were grown under greenhouse conditions throughout the summer, fall and
winter seasons of 2015 through 2017. Plants were positioned under high-pressure sodium lamps;
either 250-watt, Hydrofarm Horticultural Products model SBS250 or 24-watt Hydrofarm
Horticultural Products model F24T5, approximately 0.61 m from the lights resulting in 19.37
µmol/m2/sec and 11.70 µmol/m2/sec incident light, respectively. Plants were exposed to 18:6 light/dark cycles (slightly longer than natural day length to promote quicker and more abundant
growth) along with light coming in from the greenhouse windows (Brown 2017; Park & Runkle
2017). Temperatures fluctuated within the greenhouse according to season in which summer
maximum temperatures reached 35 ⁰C in the summer and minimum temperature was 5 ⁰C in the winter.
Plants were irrigated with deionized water, receiving 475 to 590 mL daily or more
depending on season and need to keep soil moisture at 70%. Neutral compounds (like CBZ and
other plant nutrients) are taken up in plants by diffusion through plant plasmodesmatas (Bartrons
& Peñuelas 2017) and so it was important to keep the tomato plants grown in this study at a
constant rate of watering. Also, to compensate for nutrient deficiency, the plants were given
20
Figure 6. The hutches that were built to house individual tomato plants, hornworms and moths.
Planting matrix consisted of a mixture of 50/50 commercial topsoil (Mountain Country
Topsoil, Permagreen Organics Co, Arvada, CO) and Sphagnum spp. peat moss (Fertilome All
Natural Organic Pure Canadian Sphagnum Peat Moss). An average of 1.25 +/- 0.06 kg of
planting matrix was added to each pot. Soil spike was completed for each pot individually so that
the variation of amount of soil in pots would not be a factor of CBZ transfer from soil to plant.
All pots received the same amount of CBZ.
Tomato seeds were sown in 3.5 L plastic growing pots (except for the third set which
were grown in terra-cotta pots of the same volume). All the seeds were allowed to germinate and
then pruned arbitrarily down to one plant per pot once the first true leaves began to show. Pots
21
filled with gardening soil and the soil inside the pots was separated from the dirt in the bottom of
the hutches by plastic water catchers (commonly used for indoor plants). Tomatoes were allowed
to grow to full maturity (approximately 130 days) or until there was enough foliage to support
multiple hornworms. At approximately 7 weeks of growth the apical meristems of all tomatoes
were removed to allow for increased growth of leaves and stems already present and for the
secondary meristems to bud and grow (Traas & Doonan 2001). This was done to ensure that
there was enough plant material for the hornworms to feed on. Plant samples were harvested on
the day that the hornworms were added to the plants, approximately 2 to 5 grams wet weight.
Leaf tissue only was obtained from all areas around the plant, both new and old leaves to ensure
a complete sampling of the plants.
Rearing Hornworms and Moths
Hornworm eggs were obtained from Carolina Biological. Eggs hatched in their travel
tubes after 3 days of remaining in a warm place. 3 to 8 hornworms were placed on each
individual tomato plant and allowed to grow either to their final instar before pupation (fifth
instar) or to full adult moth. Hornworms experienced high mortality (50%) but 1 to 2 hornworms
or hawkmoths survived to each of the growth limit stages for most plants. Hornworms fed on
tomato leaves of control plants, carrier control plants and treatment plants and their excrement
was also collected, separated by treatment group. The hornworms that went to the moth stage
were allowed to burrow in the dirt at the bottom of the hutch. Dirt was used as a burrowing
matrix to imitate an environment that hornworms would naturally experience (Joesten et al.
22
procured within 24 hours of emergence. During the pupation period the soil was kept moist
(60%) to ensure moth survival.
Sample Analysis
Analytical methods were developed for tomato and hornworm tissue. A standard
calibration curve was made using previously determined (unpublished) methods using
pressurized liquid extract (PLE) and analysis by UHPLC/MS/MS. In extraction sample cleanup
was used. Three different sorbents were tested (Alumina, Silica and Florisil) for performance
during PLE to determine which reduced matrix in the final plant tissue extract samples (which
can cause excessive noise and peak broadening/-tailing during instrumental analysis). Once the
absorbent alumina (Aluminum Oxide, Al) was selected, various ratios of Al to plant material
were tested along with testing additional sample preparation parameters, such as sample
concentration by evaporation or none evaporation of sample extracts simultaneously. Using a 1
to 5 plant to sorbent ratio with no evaporation gave the best instrument response with regards to
peak quality and analyte recovery for plant samples. During method validation all of the test
plant samples were spiked with regular carbamazepine and isotope labeled carbamazepine (13C-6
from Cerilliant; 100 µg/mL in 1 mL MeOH) to a final concentration of 5 ng/mL and 1 ng/mL respectively.
Silica was initially chosen as an absorbent for extraction because it is a commonly used
sorbent for matrix components expected to be a problem in the hornworm samples such as lipids
(Simon et al. 2015). Varying ratios of silica to hornworm sample were tested with a ratio of 1 to
3 worm sample to sorbent by mass yielding the best results. Worm samples were also tested for
23
sample concentration will increase the instrument response to the analyte(s) in a sample it can
also magnify any matrix effects as well. Ultimately extract concentration by evaporation (70 ⁰C with constant nitrogen flow) resulted in greater sensitivity for the analysis. Whole extracts
(approximately 35 mL) were evaporated down to 1.5 mL and then filtered into LC vials using a
syringe filter (2.5 cm filter with 0.2 µm PTFE Membrane). For method validation all of the hornworm samples were again spiked with labeled and unlabeled CBZ to attain a final
concentration of 1ng/mL and 5 ng/mL respectively. Hawk moths were treated in the same
manner as the worm samples. However a 12 to1 sorbent to moth ratio was implemented because
on instrumental analysis this type of sample prep gave the sharpest peak and best response for
expected concentration of CBZ spiked into the test sample. Tests were also performed by leaving
the moth whole and comparing the differences in response to moths that had been chopped prior
to extraction. The results suggested the benefits of cutting up the hawk moths prior to extraction
based on analyte recovery.
Extractions
Pressurized liquid extraction (PLE) using a Dionex ASE-100 was used to extract
carbamazepine from plant, hornworm and hawkmoth tissue (Figure 7). The extraction solvent
was a 1:1:1 mixture of methanol (MeOH), acetonitrile (ACN) and dichloromethane (DCM).
24
Figure 7. The ASE instrument that was used to complete all of the sample extractions in this study.
compounds. DCM is also an organic solvent that has the ability to be used to extract the above
mentioned compounds however it is not as commonly used (Brooks et al. 1998). After samples
were collected at the end of their allotted growth periods they were stored at -20 ⁰C until
extraction and analysis. For plant tissue samples a glass microfiber filter was placed at the
bottom of a 10 mL ASE cell. A layer (~ 1 cm) of ashed sand (400 ⁰C 10hr) was placed on top of the filter. This was followed by 5 g of alumina which was the sorbent used for plant extractions.
Approximately 1 gram of homogenized plant tissue was added to the cell. A remaining void
volume was filled with ashed sand to maintain consistent extract volumes. The PLE method
consisted of 2 static cycles, 11 min each, at 130 ⁰C and 1500 kPa.
Hornworms and hawkmoths were extracted in a similar manner as the plant samples.
However, for these two sample types, silica was used as the sorbent for in-extraction cleanup
25
matrix. Hornworms and moths were extracted using the larger ASE cells (60 mL) to
accommodate the lager sample size of the hornworms and hawkmoths. As with extraction of
plant tissue a glass fiber filter was placed in the cell first. Sand was scooped into the cell to
approximately one half of an inch. Silica was then weighed into the cell using a ratio of about 3:1
silica to worm or 12:1 silica to moth sample. Hornworms and moths were taken directly from the
freezer and coarsely chopped using a clean disposable razor blade and a clean watch glass. Most
worm and moth samples were chopped with their own razor blade as well and if not the blade
was cleaned between each sample with soap and water, rinsed with ultra-pure water and finally
acetone. Any remaining void volume was filled with ashed sand. The same extraction parameters
were used for hornworm and hawkmoth samples.
All samples were collected into 250 mL clean glass collection bottles and stored at 4 ⁰C until preparation for instrument analysis. Plant extracts were analyzed directly with minimal
additional preparation. Plant extract (1.35 mL) was pipetted into a 2 mL autosampler vial along
with 150 µL of 10 ng/mL labeled CBZ internal standard. Plant samples were prepared in
triplicate using the same plant extract solution. Once samples were in the autosampler vials they
were either analyzed immediately or again stored at 4 ⁰C for later analysis by LC/MS/MS. To prepare hornworm and hawkmoth extracts for instrumental analysis, 150 µL of 10
ng/mL isotope labeled CBZ internal standard was pipetted into the whole extract mixed. The
whole extract was quantitatively transferred into glassware for a LABCONCO RAPIDVAP N2
Evaporation system. Extract was transferred using 5 mL ACN to rinse the bottle and was added
to the extract in the evaporating cell. Extracts were evaporated down to 1.5 mL at 70 ⁰C under a
Luer-26
Figure 8. The UHPLC MS/MS that was used to analyze all of the samples qualitatively and quantitatively for carbamazepine.
Lock plastic disposable syringe with a 2.5 cm/ 0.2 µm PTFE membrane syringe filter attached.
The extract was filtered into a 2 mL autosampler vial and was either analyzed immediately or
was stored at 4 ⁰C for later LC/MS/MS analysis.
Two method blanks (ashed sand) were extracted with each sample set. The method
blanks were prepared in the same fashion as the tissue sample was being extracted. The method
blanks were the first and last samples to be extracted, respectively during each sample extraction
set.
Instrumental Analysis
All samples were analyzed on a Thermo Scientific Dionex UltiMate 3000 Ultra-High
Performance Liquid Chromatograph coupled with a Thermo Scientific Quantum Access triple
quadrupole Mass Spectrometr (UHPLC/MS/MS) (Figure 8). Carbamazepine was separated from
27
sub-2 µm silica particles and 100 mm length. The Mass Spectrometer was operated using an
electrospray ionization (ESI) source in the positive ionization mode.
The UHPLC portion of the instrument was controlled using Chromelean 6.8 Software
and the MS portion was controlled using Thermo XCaliber 3.1 software. During sample analysis
the column oven was maintained at 30 ⁰C and the auto-sampler module was maintained to 5 ⁰C.
Mobile phase flow rate was set to 0.3 mL/min and the mobile phase consisted of a binary
gradient of aqueous 0.05% formic acid in water (Component A) and methanol (Component B).
The binary gradient started with 90% A for 0.5 min then slowly decreased (while B increased)
over a period of 5.5 min until B was flowing 95% for 1 min. Then mobile phase was returned to
the initial conditions at 7.1 min. The total sample analysis time was 10 min. The retention time
for CBZ was 5.3 min. Two MRM transitions were monitored for CBZ and isotope labeled CBZ
(internal standard) on tandem Mass Spectrometer (Table 2). For the unlabeled CBZ, m/z 236
194 and 236 179 were monitored, with m/z 194 being quantified. For the isotope labeled
28
quantifying ion. Ion ratios for each of the transitions for analyte and internal standard were
monitored for positive identification. For the non-labeled CBZ the ratio of m/z 194:179 was
100:10 and for the isotope labeled internal standard the ratio of m/z 200:184 was 100:5.
Calibration Standard solutions were prepared using a stock solution of CBZ (204 ng/mL).
Six standards were made from the stock solution at concentrations of 0.5 ng/mL, 0.75 ng/mL, 1
ng/mL, 5 ng/mL, 10 ng/mL and 20 ng/mL. The solutions were made in a 50/50 mixture of water
and methanol. The lowest CBZ concentration of the calibration curve was 0.5 ng/mL and no
samples were quantified if they responded with a CBZ concentration below that value. This
value represents the lowest concentration of CBZ on the standard calibration curve that was run
with each sample during instrumental analysis. Continuous Calibration Verification (CCV)
Standards were prepared in a similar manner from an independent stock solution at concentration
1 ng/mL and 10 ng/mL. The CCV standards were run with each sample set every 10th sample to
ensure that the calibration standard was valid throughout analysis of each sample set run on the
UHPLC/MS/MS. All calibration standards, CCVs, method blanks and samples had the isotope
labeled internal standard added to a final concentration of 1 ng/mL. In addition to the method
blanks for each sample set, solvent blank samples were also run, which consisted of only the
50/50 mixture of water and methanol with internal standard.
Contamination was observed approximately one fourth of the way through running
sample sets. To eliminate glassware and other lab tools as the source of contamination various
samples were analyzed on the instrument. Glassware was rinsed with solvent solution prior to
being used, just after use and again after having been washed thoroughly and if possible, ashed
29
the solvent used for the blank samples were run on the instrument as a possible sources of
contamination. Also rinses performed after extractions were tested. All of these samples
appeared to not be the source of contamination and it was determined that excess CBZ was
coming off of the column every time samples were introduced to the column no matter if they
were blanks samples or actual test samples. Because of this occurrence, the baseline used to
determine valid concentrations was set at an average of the response that the blanks were
showing.
Sample Concentration
Concentrations in whole organism tissues were determined using Equation 1.
𝐶𝐵𝑍 𝐶𝑜𝑛𝑐𝑒𝑛𝑡𝑟𝑎𝑡𝑖𝑜𝑛 𝑖𝑛 𝐿𝐶 𝑣𝑖𝑎𝑙 (𝑛𝑔 𝑚𝑙)⁄ 𝑋 𝐸𝑥𝑡𝑟𝑎𝑡 𝑉𝑜𝑙𝑢𝑚𝑒 (𝑚𝐿)
𝑡𝑖𝑠𝑠𝑢𝑒 𝑠𝑎𝑚𝑝𝑙𝑒 (𝑔) =
𝑛𝑔 𝐶𝐵𝑍
𝑔 𝑡𝑖𝑠𝑠𝑢𝑒 𝑒𝑞. 1 All samples were reported in both wet weight and dry weight values. To determine the volume of
extract, the density of the extraction solution was experimentally calculated to be 0.9569 g/mL,
which was the 1:1:1 mixture of acetonitrile, methanol and dichloromethane. The difference in the
mass of the extract receiving bottles was measured and used to calculate the volume of extract
generated.
Cleaning Protocols
All glassware used during sample preparation had been thoroughly cleaned with soap and
water, solvent rinsed with HPLC grade acetone and ashed at 400 ⁰C for 10 hours (except for any volumetric glassware, which was carried through multiple solvent washes). Between each
extraction ASE cells were cleaned Along with being washed with soap and water, ASE cells
were sonicated in HPLC grade methanol for 5 minutes followed by a thorough rise with HPLC
30
rinse cycles (flushes solvent through the entire ASE). Tissue samples were extracted in a specific
order to try and alleviate contamination within the instrument. For example, all carrier control
plants, hornworms and moths were extracted first and then their counterpart ‘treatment’ samples were extracted. Multiple ASE rinse cycles were performed between each sample extraction.
Quality Control
Quality controls for the greenhouse experiments consisted of keeping growing conditions
similar between each run (with the variability of season) and to account for seasonality; three
sets of plant/ hornworm/ moth growth were completed in different seasons to compensate for the
seasonal changes that occur in a greenhouse. The first set began mid October 2015 and ran
through March 2016 to completion. The second set ran from mid-June 2016 to the beginning of
December 2016. The third and final set ran from mid-October 2016 to mid-March 2017. There
were two control groups; a negative control and a carrier control and the third group was the
treatment group. Analytical performance was assessed by determining a limit of quantitation
along with analyte recovery experiments. Multiple blanks (solvent and method) were included
with each sample set during instrumental analysis. Continuous calibration verifications (CCVs)
were also run with each sample set during instrumental analysis to ensure the standard
calibration samples were valid, that concentrations were consistent and at the expected values.
Multiple MRM transitions were monitored for CBZ and the isotope-labeled CBZ internal
standard. The instrument was calibrated prior to analysis of each sample set. The calibration
standard curve gave a linear response and so was used to quantify all samples; sample sets each
had their own linear equation to work from using the same set of calibration standards across all
31
2016, B). According to the guidelines, MDL = (t) x (S), where t is the statistics value of a 99%
confidence level and degrees of freedom of n-1 and S is the standard deviation of 2 times the
estimated detection limit. For this analysis the estimated detection limit was 0.1 ng/mL. The
anticipated instrument reporting concentration was 0.2 ng/mL and the instrument gave an
average response of 0.19 +/-0.02 ng/mL. The MDL was calculated to be 0.05 ng/mL. Because
plant and hornworm and hawkmoth samples were prepared differently and the % moisture was
different between the three different sample types, individual detection limits (DL) were
calculated for the varying sample types. The instrument MDL (determined using EPA method)
was used to find the sample DL’s by multiplying by the appropriate extraction volume and then dividing by the average sample mass (g). DL for plant samples was 5.43 ng/g, for hornworm
samples it was 0.12 ng/g and the DL for hawkmoth samples was 0.17 ng/g. The Limit of
Quantitation (LOQ) was also determined as an approximation of the MDL and DL. This value
was applied to the data for statistical analysis. Any value that fell below LOQ was considered
zero in statistical analysis. LOQ for plant samples was 18.07 ng/g, for hornworm samples it was
0.38 ng/g and for hawkmoth samples it was 0.57 ng/g.
Statistical Analysis
A two-way ANOVA was applied to the data with CBZ treatment and organism (plant,
hornworm or hawkmoth) and their interaction as explanatory variables. Negative and carrier
controls were combined and compared to CBZ exposed organisms. All of the data was 1+ log
transformed to adjust for normality assumptions of the two-way ANOVA. All statistics were
performed in RStudio (Version 0.99.473, 2015) or Microsoft Excel (2010).A hypothesis test of
32
groups of plant and hornworm samples. The data was log transformed to fit the assumptions of
the test which accounted for the outlier data point.
Results
Statistical Analysis
The 2-Way ANOVA results showed a significant difference between treatment groups;
control and CBZ and all three organism types; plants, hornworms and hawkmoths. The F-value
between treatment types was 196.38 and the p-value was 2.20x10-16 (α= 0.05). There was also a significant difference between plant, hornworm and hawkmoth groups in general. The F-value
for this grouping was 8.94 and the p-value was 4.21x10-4 (α= 0.05). Negative and carrier control groups of plant and hornworm samples were combined for the two-way ANOVA because in both
groups there was no significant difference between the two control groups as seen by initial
statistical analysis not reported here. Also the hawkmoth group had no negative control
organisms and so by combining the negative and carrier control groups of the plant and
hornworm samples the data was consolidated and fit the 2-Way ANOVA parameters better.
Statistically significant differences were found between the control and CBZ treatment groups
for both plants and hornworms but a significant difference was not found between treatment
groups for the hawkmoth samples (Figure 9). A statistically significant difference was found
between plant samples and hornworm and hawkmoth samples but not between hornworm and
hawkmoth samples. In general there is a statistically significant difference between control and
33
Tomato Plant Data
Plant groups had 9 samples (n=9). Sample data is reported in dry weight (dw).The plants
in the negative control group did not contain detectable quantities of CBZ; not detected (ND). As
a result there was no numerical value to associate with this group and so in statistical analysis the
negative control plants were assigned a value of zero. One sample of the carrier control plant
group samples had no detectable CBZ in it, four of the samples were below the detection limit (<
DL) and four samples gave an instrumental response that was below the limit of quantitation (<
LOQ). No numerical value is reported for samples that give results below the LOQ, DL or are
ND so zero was assigned to these samples for statistical analysis. The negative control and
carrier control groups do not have averages or standard deviations associated with them (Table Figure 9. This graph displays back transformed concentrations (dw) for plant, hornworm and hawkmoth samples for control and CBZ treatment groups. Negative and carrier control group samples were combined for statistical analysis. (*) indicates statistical differences between treatment groups. (**) indicates statistical differences between organism groups.
-20 0 20 40 60 80 100 120 140
Plant Hornworm Hawkmoth
C B Z C oncent rat ion (ng/g) C T * ** *
34
3). All of the samples in the treatment group gave an instrument response within the calibration
range. The average concentration of CBZ in the treatment group was 184.77 +/- 103.74 ng/g
(dw). Data was skewed right and the application of a 1+ log transformation was used to improve
the fit of normal distribution. Two-way ANOVA statistical analysis was performed on the log
transformed data and showed a statistically significant difference between the control group plant
samples and the CBZ treatment group plant samples (Figure 9). The CBZ treatment group has a
much higher concentration than both the negative and carrier control groups (Figure 10).
Table 3. Plant Samples for all treatment types and individual sample numbers (ng/g)
ND = Not Detected, <LOQ = below Limit of Quantitation, <DL = below sample detection limit, N/A no value available; concentrations reported in dry weight (dw)
35 0 50 100 150 200 250 NC CC T C B Z C oncent rat ion (ng/g) Treatment Type Hornworm Data
Sample size varied among treatment groups for the worms because of loss during
sample preparation and because of hornworm mortality. The negative control group had a total
of seven samples (n=7). The carrier control group had nine samples (n=9) and the CBZ treatment
group had eight samples (n=8). The average CBZ concentration in negative control worms was
4.65 +/-7.90 ng/g (dw). During sample analysis of worms, instrumental contamination occurred
and so this group has a higher concentration value than the carrier control group (Figure 11). The
average CBZ concentration for the carrier control group was 2.19 +/-1.86 ng/g (dw). The average
concentration of CBZ in the hornworms that consumed plants material grown in CBZ fortified
soil 27.48 +/-32.20 ng/g (dw). The data was skewed to the right and so was normalized by
performing a 1+ log transformation. Two-way ANOVA statistical analysis, performed to Figure 10. The average CBZ concentrations (on a dry weight basis) of the plant samples for all three treatment groups and has not been manipulated for statistical analysis. NC is the negative control group. CC is the carrier control group and T is the CBZ treatment group. Error bars represent standard error of the mean.
36 0 5 10 15 20 25 30 35 40 45 NC CC T C B Z C oncent rat ion (ng/g) Treatment Type
compare the treatment groups, showed a statistically significant difference between control and
CBZ treatment hornworm samples (Figure 9). Among samples the CBZ treatment group has the
highest CBZ concentration per gram of sample (Table 4).
Figure 11. The average CBZ concentrations (dw) in the hornworm samples for all three treatment groups and has not been altered for statistical analysis. NC is the negative control group, CC is the carrier control group and T is the CBZ treatment group. Error bars represent standard error of the mean.
Table 4. Worm samples for all treatment types and individual sample numbers (ng/g)
No Sample= No organism analyzed, <LoQ= below limit of quantitation, samples reported in dry weight
37 Hawkmoth Data
No hawkmoth samples were acquired for the negative control group because none of the
hawkmoths survived through pupation. The carrier control and CBZ treatment groups had 6
moth samples each (n=6). In the carrier control group one sample resulted in a ND response and
one sample resulted in < LoQ response (Table 5). The average CBZ concentration in the carrier
control samples is 1.65 +/-2.20 ng/g (dw). Hawkmoths grown under CBZ treatment gave
instrument responses within the range of the calibration curve. The average CBZ concentration
for these samples is 3.76 +/-2.77 ng/g (dw). The treatment group average is higher than the
carrier control group average of CBZ (Figure 12) but not at a statistically significant level as
shown by the two-way ANOVA performed on the 1+ log transformed data (Figure 9).
Table 5. Moth samples for all treatment types and individual sample numbers (ng/g)
ND= Not Detected, No Sample= No organism analyzed, <LoQ= below limit of quantitation, N/A= no value available, concentrations reported in dry weight