Citation
Development and Phenological Modeling of the Oriental Latrine Fly, Chrysomya Megacephala (Diptera

Material Information

Title:
Development and Phenological Modeling of the Oriental Latrine Fly, Chrysomya Megacephala (Diptera Calliphoridae)
Creator:
Gruner, Susan V
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (96 p.)

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Entomology and Nematology
Committee Chair:
CAPINERA,JOHN LOWELL
Committee Co-Chair:
OI,FAITH M
Committee Members:
MCAUSLANE,HEATHER J
NICKERSON,MAX ALAN
SLONE,DANIEL H
Graduation Date:
12/19/2014

Subjects

Subjects / Keywords:
Ambient temperature ( jstor )
Average linear density ( jstor )
Eggs ( jstor )
Forensic entomology ( jstor )
Insects ( jstor )
Instars ( jstor )
Larvae ( jstor )
Liver ( jstor )
Modeling ( jstor )
Parametric models ( jstor )
Entomology and Nematology -- Dissertations, Academic -- UF
forensic
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Entomology and Nematology thesis, Ph.D.

Notes

Abstract:
Chrysomya megacephala, the oriental latrine fly, is found throughout the tropical regions of the world, with recent expansion into more temperate areas. This calliphorid is also an important tool in forensic entomology as it is one of the first species of calliphorid flies to arrive at a corpse; thus, knowledge of the temperature-dependent development is a key factor in estimating the ages of specimens collected at a scene for determine of a postmortem interval (PMI). The relationship between volume, number of maggots, and different combinations of instars on maggot mass heat generation was determined by comparing different numbers of C. megacephala larvae (40, 100, 250, 600 and 2000), and different combinations of instars (approximately 100% first instars, 50/50 first/second instars, 100% second instars, 50/50 second/third instars, and 100% third instars) at two different ambient temperature (21.2 and 30.8 C). Twelve candidate multiple regression models were fitted to the data; the models were then scored with Akaike information criterion (AIC) and ranked with the Bayesian information criterion (BIC). The results of this study, based on the BIC rank, indicate that although instar, age, treatment temperature, and number of maggots in a mass were significant, larval volume was the best single 10 predictor of maggot mass temperatures, but the most parsimonious model incorporated larval volume and ambient temperature. The accumulated development time and population transition points from oviposition for each life stage from eclosion to adult emergence were determined at five constant temperatures (nominal/actual): 15/16.1, 20/21.2, 25/ 26.1, 30/30.8, and 35/35.6 C. For each transition, the 10th, 50th and 90th percentiles were calculated with a logistic linear model. The mean population transition times and % survivorship were determined directly from the raw laboratory data. Linear and curvilinear phenological models were compared for best fit with the C. megacephala development times. The curvilinear models predicted development rates better than the linear models. Although linear models were effective using intermediate temperature, maggot mass temperatures frequently exceeded the upper developmental thresholds. The linear models were less effective under these conditions, but curvilinear models improved predictive capabilities. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis:
Thesis (Ph.D.)--University of Florida, 2014.
Local:
Adviser: CAPINERA,JOHN LOWELL.
Local:
Co-adviser: OI,FAITH M.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2015-12-31
Statement of Responsibility:
by Susan V Gruner.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Embargo Date:
12/31/2015
Resource Identifier:
974372121 ( OCLC )
Classification:
LD1780 2014 ( lcc )

Downloads

This item has the following downloads:


Full Text

PAGE 1

1 DEVELOPMENT AND PHENOLOGICAL MODELING OF THE ORIENTAL LATRINE FLY, CHRYSOMYA MEGACEPHALA (DIPTERA: CALLIPHORIDAE) By SUSAN V. GRUNER A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2014

PAGE 2

2 © 2014 Susan V. Gruner

PAGE 3

3 To my mother, Rosamo nd, and to my husband, Michael

PAGE 4

4 ACKNOWLEDGMENTS I extend my deepest gratitude and love to my husband, Michael, without whom I could not have accomplished th is research. There are not too many spouses who would put up with smelly liver , slimy maggots and a sleep deprived wife, but Michael did, with little to no complaints. I am eternally grateful to my mother, Rosamond, for her love, encouragement and support. I extend my deepest gratitude to my supervisory chairman, Dr. John Capinera, for his guidance, assistance and support in this research . Dr. Capinera was probably the only one in the department who did not to complain about the smell. Thank you to Dr. Daniel Slone , for his statistical expertise , guidance , friendship and en thusiasm for this research. Appreciation is expressed to Dr. Max Nickerson, Dr. Faith Oi, and Dr. Heather McAuslane for serving on my committee. I would also like to thank Dr. Kenneth Schoenly, Dr. Leon Higley, Dr. Neal Haskell and Dr. Jeff rey Wells for th eir kind support and advice. There were others in the entomology department who were always kind and helpful to me, especially Debbie Hall and Nancy Sanders.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ .......... 9 ABSTRACT ................................ ................................ ................................ ................... 10 CHAPTER 1 LITERATURE REVIEW ................................ ................................ .......................... 12 The Calliphoridae ................................ ................................ ................................ .... 12 Impor tance of Calliphorids as Decomposers ................................ .......................... 12 Forensic Importance of Calliphorids ................................ ................................ ....... 13 Chrysomya megacephala (Fabricius) ................................ ................................ ..... 15 Feeding Substrate for Calliphorid Larval Development ................................ ........... 15 Volume of Larvae as an Indicator of Maggot Mass Temperature ........................... 16 Development of Chrysomya megacephala at Constant Temperatures ................... 16 Linear and Nonlinear Modeling of Chrysomya megacephala Development Rates ................................ ................................ ................................ ................... 17 Linear Model (Degree Day Model) ................................ ................................ .......... 17 The Modified Sharpe and DeMichele Model (Schoolfield et al. 1981) .................... 19 The Modified Logan et al. Model (Lactin et al. 1995) ................................ .............. 20 2 A FRESH LIVER AGAR SUBSTRATE FOR REARING SMALL NUMBERS OF FORENSICALLY IMPORTANT BLOW FLIES (DIPTERA: CALLIPHORID AE) ....... 24 Introduction ................................ ................................ ................................ ............. 24 Materials and Methods ................................ ................................ ............................ 25 Results and Discussion ................................ ................................ ........................... 27 3 VOLUME OF LARVAE IS THE MOST IMPORTANT SINGLE PREDICTOR OF MASS TEMPERATURES IN THE FORENSICALLY IMPORTANT CALLIPHORID CHRYSOMYA MEGACEPHALA (DIPTERA: CALLIPHORIDAE) .. 29 Introduction ................................ ................................ ................................ ............. 29 Methods and Materials ................................ ................................ ............................ 31 Insect Culture ................................ ................................ ................................ ... 31 Effect of Number of Larvae in a Mass ................................ .............................. 33 E ffect of Larval Instar in a Mass ................................ ................................ ....... 33 Effect of Larval Volume in a Mass ................................ ................................ .... 33 Environmental Chamber Temperature Profile ................................ .................. 34 Statistical Analyses ................................ ................................ .......................... 34

PAGE 6

6 Results ................................ ................................ ................................ .................... 36 Environmental Chamber Temperatures ................................ ............................ 36 Number of Larvae in a Mass ................................ ................................ ............ 36 Larval Inst ars in a Mass ................................ ................................ .................... 36 Maggot Mass Volume ................................ ................................ ....................... 37 Predictive Models of Maggot Mass Temperature ................................ ............. 37 Discussion ................................ ................................ ................................ .............. 38 4 DEVELOPMENT OF THE ORIENTAL LATRINE FLY, CHRYSOMYA MEGACEPHALA (FABRICIUS) (DIPTERA: CALLIPHORIDAE) AT FIVE CONSTANT TEMPERATURES ................................ ................................ .............. 45 Introduction ................................ ................................ ................................ ............. 45 Materials and Methods ................................ ................................ ............................ 48 Environmental Chamber Temperature Profile ................................ .................. 49 Egg Collection ................................ ................................ ................................ .. 50 Laboratory Development Time Tests ................................ ............................... 50 Field Data Collection ................................ ................................ ........................ 52 Statistical Analyses ................................ ................................ .......................... 52 Results ................................ ................................ ................................ .................... 53 Environmental Chambers ................................ ................................ ................. 53 Laboratory Development without Effect of Maggot Mass ................................ . 54 Field Trials with Effect of Maggot Mass Temperatures ................................ ..... 55 Discussion ................................ ................................ ................................ .............. 55 Importance of Ambient and Maggot Mass Temperatures for PMI Estimations ................................ ................................ ................................ .... 57 5 COMPARISON OF THREE PHENOLOGICAL MODELS FOR BEST FIT OF DEVELOPMENT TIMES FOR THE ORIENTAL LATRINE FLY, CHRYSOMYA MEGACEPHALA (DIPTERA: CALLIPHORIDAE) ................................ ................... 68 Introduction ................................ ................................ ................................ ............. 68 Linear Model (Degree Day Model) ................................ ................................ ... 69 The M odified Sharpe and DeMichele Model (Schoolfield et al. 1981) .............. 70 The Modified Logan et al. Model (Lactin et al. 1995) ................................ ........ 71 Materials and Methods ................................ ................................ ............................ 72 Results ................................ ................................ ................................ .................... 73 Discussion ................................ ................................ ................................ .............. 73 6 CONCLUSIONS ................................ ................................ ................................ ..... 78 APPENDIX A: R CODES FOR CHAPTER 3 ................................ ................................ .................... 81 B: TIMES TO TRANSITION (H) AT THE 10TH , 50TH, AND 90TH PERCENTILE ....... 82 C: R CODE FOR DEVELOPMENTAL POPULATION TRANSITION TIMES ................ 84

PAGE 7

7 D: R CODES FOR MODEL PARAMETER ESTIMATION ................................ ............. 85 LIST OF REFERENCES ................................ ................................ ............................... 86 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 96

PAGE 8

8 LIST OF TABLES Table page 1 1 Sample temperatures in °C, development time (T), and development rate (1/T) (adapted from Sharov 1998) ................................ ................................ ...... 22 1 2 Sample days, mean temperatures (t), effective temperature (t x min ) accum ulated degree days (ADD). ................................ ................................ ...... 23 2 1 Survivorship (percent 1 st instar larva to adult eclosion) of Chrysomya megacephala (F.) on fresh liver and liver agar substrates. ................................ . 28 3 1 Twelve candidate models and their AIC and BIC scores. The model with the l owest score is considered to be the best. ................................ .......................... 44 4 1 Cumulative times from oviposition to each life stage transition for Chrysomya me gacephala at 16.0 °C constant temperature ................................ .................. 64 4 2 Cumulative times from oviposition to each life stage transition for Chrysomya me gacephala at 21.2 °C constant temperature. ................................ ................. 64 4 3 Cumulative times from oviposition to each life stage transition for Chrysomya megacephala at 25.8 °C constant temperature. ................................ ................. 65 4 4 Cumulative times from oviposition to each life stage transition for Chrysomya megacephala at 30.8 °C constant temperature. ................................ ................. 65 4 5 Cumulative times from oviposition to each life stage transition for Chrysomya megacephala at 35.6 °C constant temperature ................................ .................. 65 4 6 Predicted error in development time (h) from oviposition to adult emergence, per 1.0 °C error in environmental chamber temperature. ................................ ... 66 4 7 Mean and peak larval mass temperatures from pig carcasses in field trials conducted in Earleton, FL, from 2002 2004. ................................ ...................... 67 5 1 Seven development models and their estimated parameters values. A lower RSS indicates a better fit. ................................ ................................ ................... 77

PAGE 9

9 LIST OF FIGURE S Figure page 1 1 Linear regression of development rate as a function of temperature with the sample data from Table 1 1. ................................ ................................ ............... 22 3 1 Larval mass temperatures as a function of number of larvae in a mass at two different treatmen t temperatures (21.2 and 30.8 °C) ................................ .......... 41 3 2 Larval mass temperatures as a function of larval instar at two different ambient tempera tures (21.2 and 30.8 °C) ................................ .......................... 42 3 3 Larval mass temperature as a function of volume. Larval masses with greater volumes had higher med ian temperatures ................................ ......................... 43 4 1 Cup arrangement (Photo courtesy of author) ................................ ..................... 60 4 2 Ribbon graph depicting development stages and transition times of C. megacephala compared to that of other studies of C. megacephala . ................. 61 4 3 Ribbon graph depicting development times of C. megacephala compared to development from published studies of C. rufifacies ................................ .......... 62 4 4 Ribbon graph depicting our development times of C. megacephala compared to development from published studies of L. sericata ................................ ......... 63 5 1 Chrysomya megacephala development rates fitted in models 1,4,5,6, and 7. .... 76

PAGE 10

10 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy DEVELOPMENT AND PHENOLOGICAL MODELING OF THE ORIENTAL LATRINE FLY, CHRYSOMYA MEGACEPHALA (DIPTERA: CALLIPHORIDAE) By Susan Victoria Gruner December 2014 Chair: John Capinera Major: Entomology and Nematology Chrysomya megacephala (Fabricius) , the orie ntal latrine fly, is found throughout the tropical regions of the world, with recent expansion into more temperate areas . This calliphorid is also an important tool in forensic entomology as it is one of the first species of calliphorid flies to arrive at a corpse ; thus, knowledge of the temperature dependent development is a key factor in estimating the ages of specimens collected at a scene for determine of a post mortem interval (PMI). T he relationship between volume, number of maggots, and different com binations of instars on maggot mass heat generation was determined by comparing different numbers of C . megacephala larvae (40, 100, 250, 600 and 2000), and different combinations of instars (approximately 100% first instars, 50/50 first/second instars, 100% second instars, 50/50 second/third instars, and 100% third instars ) at two different ambient temperature (21.2 and 30.8 °C). T welve candidate multiple regression models were fitted to the data; the models were then scored with Akaike information crite rion (AIC) and ranked with the Bayesian information criterion (BIC). The results of this study indicate that although instar, age, treatment temperature, and number of maggots in a mass were significant, larval volume was the best single predictor of maggo t mass

PAGE 11

11 temperatures , but the most parsimonious model inc luded larval volume and treatment temperature. T he accumulated development time and population transition points from oviposition for each life stage from eclosion to adult emergence were determined at five constant temperatures (nominal/actual): 15 /16.1 , 20 /21.2 , 25 / 26.1 , 30 /30.8, and 35/35.6 ° C . For each transition, the 10 th , 50 th and 90 th percentiles were calculated with a logistic linear model. The mean population transition times and % survivors hip were determined directly from the raw laboratory data. Small amounts (10 g or less) of fresh or frozen liver desiccated in a few hours ; thus , a liver agar diet for rearing small numbers of calliphorid maggots was developed . Linear and nonlinear phenological models were compared for best fit with the C. megacephala d evelopment times . Although linear models were effective using intermediate temperature s, but maggot mass temperatures frequently exceed intermediate temperatures. Nonlinear models imp roved predictive capabilities.

PAGE 12

12 CHAPTER 1 LITE R ATURE REVIEW The Calliphoridae The Calliphoridae, or blow flies, are synanthropic and heliotropic. They are considered to be pests because some can mechanically transmit disease causing organisms that affect humans (James 1947, Hall 1 948) and animals (Zumpt 1965). Calliphorids also can be beneficial. They are pollinators of some species of orchids (Lehneback and Robertson 2004), alfalfa, the dead horse arum ( Helicodiceros muscivorous (L.f.) Engler ), carrots, onions, and turnips, and they can increase crop seed yields (Schittenhelm et al. 1997, Bradshaw et al. 2002). Calliphorid larvae can be used for therapeutic purposes. Maggot therapy or maggot debridement therapy is used to treat chronic wounds such as osteomyelitis and venous ulcers (Nigam et al. 2006, Gruner 2008 , Gilead et al. 2012 ). Larval secretions have antibacterial activity and can inhibit growth of bacteria such as Gram positive Staphylococcus aureus and S. pyogenes (Kerridge et al. 200 5). Importance of Calliphorids as Decomposers The Calliphoridae are sarcosaprophagous insects (sarco = skin; sapro = rotten); their role as decomposers of animal and plant litter, animal feces and carrion ensures the quick return to the ecosystem of organi c materials in organic debris (Putman 1978). It is only through the decay of carrion and dung that nutrients sequestered by the animal component of any ecological community can be returned to the system for re use (Putman 1978). Calliphorid larvae are the most significant group of arthropods associated with carrion; the larvae consume more than 80% of available carrion materials during their growth and development (Putman 1978, 1983 ). Calliphorid larvae

PAGE 13

13 hasten liquefaction of muscle tissue as they feed, res ulting in a more favorab le medium for activity of micro organisms. As the maggots tunnel within a carcass, they increase the aeration of the carrion mass, enabling aerobic bacteria to penetrate deeper into the corpse (Putman 1978 ). Payne (1965) compared dec ay rates of pig carrion during the dependen Decomposing carrion , free of insects , dried slowly. After 100 days, 20% of the original carri on r emained in mummified form. During the summer months, the carcasses exposed to insects were reduced to dried skin and bones in less than a week. Forensic Importance of Calliphorids Another important benefit of calliphorids is their predictable presence on corpses. Their presence can enable i nvestigators to estimate a post mortem interval or (PMI) (Nuorteva et al. 1967, Easton and Smith 1970, Lane 1975, Nuorteva 1977, Putman 1978, Keh 1985). Successional studies are important because they establish what calli phorid species arrive at different stages of the decomposition process. Calliphorid succession varies by season, geographical location, and by carrion type (Anderson 2001a). Forensic entomology is the most accurate method of determining a PMI when more tha n a day or two have elapsed (Kashyap and Pillay 1989). Additionally, forensic entomology might be used to determine whether a body has been moved from one location to another, whether a body has been disturbed after death, and can indicate the position and presence of wounds (Anderson 2001b). Because human cadavers are not easily obtainable for decomposition studies, successional studies have been conducted using non human animal models such as pigs, dogs, cats, and mice.

PAGE 14

14 Using insects (usually calliphorid larvae) collected at death scenes, forensic entomologists currently use published development rates and the degree day model for determination of a time since death, or post mortem interval (PMI) (Catts and Haskell 1990). Larval age estimates can be imprec ise for older development stages for a number of reasons. For example, size alone should not be used as an indicator of age determination because there may be variations resulting from food availability among cohorts. Smaller larvae would be assumed to be younger, which would result in an underestimate of elapsed time since death (Anderson 2000). To determine a PMI, development rates as a function of temperature are needed for forensically important calliphorids. There are development rate studies for othe r forensically important calliphorids such as Phormia regina ( Meigen ) (Kamal 1958, Nishida 1984, Greenberg 1991, Introna et al. 1991, Anderson 2000, Byrd and Allen 2001), Calliphora vicina ( Robineau Desvo idy) (Kamal 1958, Greenberg 1991 , Introna et al. 1991, Anderson 2000, Ames and Turner 2003, Donovan et al. 2006), Chrysomya rufifacies (Macquart) (Greenb erg 1991, Byrd and Butler 1997) , Cochliomyia macellaria (Fabricius) (Greenberg 1991, Byrd and Butler 1997) and Lucilia sericata (Meigen) (Kamal 1958, Greenberg 1991, Anderson 2000, Grassberger and Reiter 2001) but there is no standardized method of obtaining development rates or times. Variables such as ambient temperatures, diet and substrate, humidity, photoperiod, constant vs. fluctuating tempe eggs or larvae) differ among the various studies. A major problem with most of the published development rate studies is that they lack replication.

PAGE 15

15 Chrysomya megacephala (Fabricius) Ch rysomya megacephala , the oriental latrine fly, is found throughout the tropical regions of the world, with recent expansion into more temperate areas (Greenberg 1988, Wells 1991, Martinez Sanchez et al. 2001, Tomberlin et al. 2001) . It is a filth fly that can cause secondary myiasis in humans and animals (Zumpt 1965, Sukontason et al. 2005). It also can transmit enteric pathogens and parasites, and is the dominant insect vector of helminth eggs in Malaysia (Sulaiman et al. 1988, 1989). This blow fly is also an important tool in forensic entomology as it is one of the first species of calliphorid flies to arrive at a corpse. In north central Florida, Lucilia coeruleiviridis (Macq uart) was formerly the first calliphorid to arrive at a corpse and was the dominant calliphorid found in maggot masses (Gruner et al. 2007), but C . megacephala has displaced L . c oeruleiviridi s as the dominant calliphorid fly species found in the spring, summer and fall months (personal observation). For regions in which C. megacephala is present, knowledge of the temperature dependent development is a key factor in estimating the ages of specimens collected at a scene. F rom this age, estima tion a post mortem interval (PMI) can be inferred. Feeding Substrate for Calliphorid Larval Development The most common feeding substrate for rearing calliphorid larvae is fresh or previously frozen beef liver (Kamal 1958, Anderson 2000, Grassberger and Re iter 2001, Ames and Turner 2003, Nabity et al. 2006, Richards and Villet 200 8 ). One of the problems encountered with use of liver (or any animal tissue) is that it can desiccate quickly (Kamal 1958), especially when using small amounts of 20 g or less. To determine development rates without the effect of larval thermoregulation (Slone an d Gruner 2007) fewer than 20 larvae per subsample were used . To allow for a longer time

PAGE 16

16 before desiccation effects occurred with the substrate, a mixture of liver and agar w as developed. Liver agar diets for feeding large numbers of sterile medical maggots have been previously described by Sherman and My Tien Tran (1995), but the sterile diet requires 5 days to prepare, has numerous components and is labor intensive. A simpli fied diet for rearing small numbers of calliphorid maggots was developed for research purposes (Chapter 2) . Volume of Larvae as an Indicator of Maggot Mass Temperature C harabidze et al. (2011) , Heaton et al. (2014), Rivers et al. (20 11 ) and Joy et al. (20 06) determined that elevated maggot mass temperatures were determined by larval stage (by instar) and number of individuals in a mass, with second instars not able to produce as much heat as third instar larvae. But in our field studies conducted in 2006 (Slone and Gruner 2007), it was determined that mass volume (> 20 cm 3 ) was a predictor of maggot mass temperature, and that age or larval instar had no effect, after taking volume into account, on maggot mass temperatures. However, the aggregations used in the analyses consisted mainly of combinations of instars (1 2, 1 2 3, and 2 3). The objective of this laboratory study in Chapter 3 ; therefore, was to determine if maggot mass volume was the main predict or maggot mass temperatures. Development of Chrysomya megacephala at Constant Temperatures To determine development times of C. megacephala without the effect of larval heat generation, masses consisting of only 10 larvae were used , thereby excluding this confounding factor. The objective of this st udy was to obtain d evelopment rates at eight different temperatures: 5, 10, 15, 20, 25, 30, 35, and 40 °C.

PAGE 17

17 Linear and Nonlinear Modeling of Chrysomya megacephala Development Rates According to Higley and Haskell (2001) , there are generally two approaches for measuring insect development using mathematical models; linear and nonlinear . These models try to predict mathematically the rate of development of a particular species based mainly on temperature. Linear Model (Degree Day M odel) Originally, degree day models were useful for understanding insect and plant phenology (Higley et al. 1986) in plant science, pest management and ecology. For example, the degree da y model might help a farmer calculate the best timing for pesticide applications. According to Wagner et al. (1984), the degree day concept is based on the work of several authors dating back to the mid 1800s. The degree day model assumes that a fixed qu antity of metabolic activity , under the control of time and temperature , is required to complete an insect dev elopment. According to the model , the amount of heat required to complete development does not vary, nor will it change due to fluctuating or cons tant temperatures. The model assumptions are: 1. A constant increase in development rate with increasing temperature. 2. No development occurs below the minimum developmental threshold. 3. The development rate changes at or above the maximum threshold , though ther e are different methods of treating this change . To understand the model, it is important to understand its components. Rate of development can be measured by a reciprocal of the number of days (or other time units) required for development completion (Wells and LaMotte 2001) . If an insect takes 1 0 days to reach adulthood from oviposition at a particular ambient temperature , the

PAGE 18

18 rate of development ( y ) would be 1/ 10 per day or 0. 10 per day. The physiolog ical time scale, incorporating both temperature and time is expressed in degree days or degree hours. Development rates can be calculated for an entire life cycle or for a specific stage of an organism. If we use sample temperature and development data from Table 1 1 , one would get a regression equation of a straight line , y = m x + b (model assumption 1) where b is the y intercept, m x minimum threshold temperature , defined as the temperature at which the development rate is 0 (model assumption 2) . A plot of these data shows that the development rate from 35 °C is outside the linear range, and thus must be excluded from the model (Figure 1 1) . The linear regression equation from the remaining ex ample data is y = 0. 0014 2 x min 0.0104 . Determination of developmental thresholds and degree day accumulations over time are prerequisites for use of the degree day model . Estimation of the developmental threshold temperature (the x intercept) is determined by extrapolating back the straight line plot of the daily development rate against temperature to the point where it reaches zero (Sharov 1998) . The regression equation for the accumu lated degree days ( ADD ) with becomes 0 = m x min + b ; thus, one can rewrite the equation to be If development in days is y = 1/ T and the effective temperature (daily mean temperature minus developmental threshold temperature min ) is ( t x min ) , then degree day s can be calculated with the following equation:

PAGE 19

19 which reduces to: In the example above, K = = 704 degree day s. Degree day s can be used to predict development time in cases with variable temperatures as shown in Table 1 2 . When the sum of accumulated days reaches K , development time has been completed. For example , using the daily temperature data from Table 1 2 and the param eters calculated above, one can estimate that on day 56, development has been c ompleted. The Modified Sharpe and DeMichele Model (Schoolfield et al. 1981) A physiological model of poikilotherm development was developed by Sharpe and DeMiche le (1977). The temperature dependent rate model is based upon highly complex equations by Eyring (1935) , Johnson and Lewin (1946) and Hultin (1955) . The Sharpe and DeMichele (1975) model has the following consequences: 1. It demonstrates the validity of the linear approximation (degree day concept) in the mid temperature region for some organisms. 2. It effectively establishes a low temperature threshold for development 3. It reduces the rate of development at higher temperatures, thereby establishing both an optimum and upper threshold for development. The assumptions of the Sharpe and DeMichele (1975) model are: 1. Development is regulated by a single control enzyme whose r eaction rate determines the development rate of the organism. 2. The development rate is proportional to the product of the concentration of the active enzymes and their rate constant (which in itself is temperature dependent). 3. The control enzyme can exist in two temperature dependent inactivation states as well as an active state.

PAGE 20

20 Sharpe and DeMichele (1975) identified thermodynamic constants that characterized many , and then tested the applicability of the model w ith insects: Pseudatomoscelis seriatus Reuter (cotton fleahopper) and Drosophila melanogaster Meigen (fruit fly) . The model fit the development rates well . But the Sharpe and DeMichele mo del was not well suited for non linear regression and was subsequently modified by Schoolfield et al. (1981) . The modification alleviated the non lin ear regression problem and allowed more convenient initial parameter estimation . F or cases with limited data availability at h igher temperatures , a reduced four parameter model was proposed (equation 6 in Schoolfield et al. 1981) : where r (T) is the mean development rate at temperature T (time 1 ), T is temperature in Kelvin (298 k = 25 °C), R is the universal gas constant (1.987 cal deg 1 mol 1 ), (25 °C) is the development rate at 25 °C, assuming no enzyme inactivation (time 1 ), is the enthalpy of activation of the reaction that is catalyzed by the enzyme, is the temperature (K) at which the enzyme is ½ active and ½ low temperature inactive , and is the change in enthalpy associated with low temperature inactivation of the enzyme (cal mol 1 ) (Schoolfield et al. 1981). The Modified L ogan et al. Model (Lactin et al. 1995) L ogan et al. (1976) developed a nonlinear model to describe g rowth rates of arthropods to address the shortcomings of the degree day model outside of the linear

PAGE 21

21 range of temperatures. L actin et al. (1995) modified the Logan et al. (1976) model by removing some parameters to make it easier to use. The objective was to find an appropriate model that could predict the wide ranges of insect development, not just those in the linear part. The L actin modification of the Logan et al. (1976) model is : max (T max where r (T) is development rate at temperature T, can be interpreted as a composite Q 10 value for critical enzyme catalyzed, biochemical reactions, T max is the thermal maximum, is the range of temperatures between T max and the temperature at which r al. 1976, Lactin et al. 1981). The objective of th e study in Chap ter 5 was to compare the l inear (degree day) and nonlinear models (Logan et al. 1976, Sharpe and De Michele 1977, Schoolfield et al. 1981, Lactin et al. 1995 ) using the C. megacephala development data to determine which model had the best fit.

PAGE 22

22 T able 1 1. Sample temperatures in °C, development time (T), and development rate (1/T) (adapted from Sharov 1998) Temperature, t (°C) Development time, T (D ays ) Development rate y = 1/T 5 10 200 0.005 15 100 0.010 20 60 0.017 25 40 0.025 30 30 0.033 35 35 0.029 Figure 1 1. Linear regression of development rate as a function of temperature with the sample data from Table 1 1. The 35 °C temperature is not in the linear part of the line. y = 0.00142t 0.0104 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0 5 10 15 20 25 30 35 Development Rate (1/T) Temperature ( C)

PAGE 23

23 Table 1 2. Sample days, mean temperatures (t), effective temperature (t x min ) accumulated degree days (ADD). Day Number Mean t (°C) ET (t x min ) ADD 1 15 7.7 7.7 2 18 10.7 18.4 3 25 17.7 36.1 4 23 15.7 51.8 5 24 16.7 68.5 52 17 9.7 670.9 53 15 7.7 678.6 54 18 10.7 689.3 55 15 7.7 697.0 56 22 14.7 711.7

PAGE 24

24 1 CHAPTER 2 A FRESH LIVER AGAR SUBSTRATE FOR REARING SMALL NUMBERS OF FORENSICALLY IMPORTANT BLOW FLIES (DIPTERA: CALLIPHORIDAE) Introduction T o determine a post mortem interval (PMI), it is usually necessary to know the development rates for different species of calliphorids, the primary colonizers of a corpse (Anderson and VanLaerhoven 1996, Nabity et al. 20 06, Richards and Villet 2009). Development rates for some forensically important species of c alliphorids have been examined in lab studies. The most common feeding substrate for rearing calliphorid larvae is fresh or previously frozen beef liver (Kamal 1958, Anderson 2000, Grassberger and Reiter 2001, Ames and Turner 2003, Nabity et al. 2006, Rich ards and Villet 2009). One of the problems encountered with use of liver (or any animal tissue) is that it can desiccate quickly (Kamal 1958), especially when using small amounts of 20 g or less. To determine development rates without the effect of larval thermoregulation (Slone an d Gruner 2007) fewer than 20 larvae per subsample were used . To allow for a longer time before desiccation effects occurred with the substrate, a mixture of liver and agar was developed. Liver agar diets for feeding large numbers of sterile medical maggots have been previously described by Sherman and My Tien Tran (1995), but the sterile diet requires 5 days to prepare, has numerous components and is labor intensive. Here, a simplified diet for rearing small numbers of calliphorid maggots for research purposes is described . Published in Journal of Medical Entomology, 51(3):713 715. 2014

PAGE 25

25 Materials and Methods The prepared diet consisted of 10 12 g of 150 mesh NF/FCC agar (Bio serv ® , Frenchtown, NJ), 250 mL of pureed, room temperature fresh beef liver (Newberry Cold Storage, Newberry, FL) and 250 mL of water. The liver was pureed in a blender (Black and Decker ® BL2800 Blender, Madison WI) for ~2 minutes until shiny and smooth in texture. The pur é ed liver was poured into a 0.7 L plastic container. The agar and water were placed in a 3.7 L glass b owl and mixed well with a rubber spatula. The bowl with agar/water mixture was placed in a 900 watt microwave and heated on high for 3.5 minutes, removed, stirred quickly for 15 sec, then microwaved on the high setting for another minute. Thereafter the ag ar stirred briskly for another 15 sec. The mixture was allowed to rest at room temperature for approximately 7 min until it had cooled to 60 °C. The warm agar was added to the pureed liver and mixed quickly with a rubber spatula until well blended. This mi xture hardened in about 15 minutes and was ready for immediate use. The liver agar was successfully refrigerated at 5 °C for up to 4 days. Calliphorid eggs were obtained by placing 300 g of fresh beef liver into cages of adult Chrysomya megacephala (F.) flies that were maintained in a rearing room at 28±1°C with 14:10 photoperiod and 50 65% RH. To test survival of larvae on small pieces of substrate, 10 first instar larvae were placed onto 5 g of liver agar or 5 g of fresh liver, which were in separate 10 0 mm x 15 mm polystyrene Petri dishes (Fisherbrand ® , Fish er Scientific, Pittsburg, PA). Petri dish tops were modified by burning a 3 cm diameter hole onto which organza fabric was hot glued to allow for air circulation. Set inside the 100 mm x 15 mm petri dish was a small piece of wet paper toweling that was on a 2.1 cm polystyrene Petri dish (Fisherbrand ® ) to increase humidity

PAGE 26

26 without wetting the substrate. Each Petri dish combination was placed into a 0.47 L plastic cup (Dart ® , Mason, MI) that was half fi lled with fine vermiculite. Each cup had a tight fitting lid with an 8 cm hole covered with 18x16 mesh aluminum window screen allowing air circulation. Fresh 5 g pieces of liver agar or liver was added as necessary in 12 hour intervals. As larvae approache d the wandering stage, the tops of the Petri dishes were removed to allow the larvae to move into the vermiculite to pupariate. Six cups were placed into a 27 L plastic container (Sterilite ® , Townsend, MA) in which a 40 x 5 cm hole had been cut on each lon g side and covered with 18x16 mesh aluminum window screen to allow for air circulation. Each Sterilite ® container was then placed on the center shelf of a Florida Reach I n environmental chamber. Two 7. 8L containers were filled with water and placed in each chamber to maintain 60% RH at all times during testing until all larvae had pupated. Th e procedure with 10 larvae on 5 g each was replicated four times in four separate chambers (two at 25 °C and two at 30 °C), all with a 14:10 photoperiod. Surviving adults were counted, sexed and added to the existing colony. To test survival of larger numbers of larvae on larger pieces of liver agar substrate, 300 first instar larvae were placed onto 50 g pieces of live r agar or 50 g of fresh liver that were in separate 1.3 L plastic containers, with tops modified with large holes covered with fine mesh for air circulation. The larvae were kept in the colony rearing room at 28±1 °C with a 14:10 photoperiod and 50 65% RH. Larvae were checked once daily and 50 g pieces of fresh su bstrate were added ad libitum. Th e procedure with 300 larvae on 50 g each of liver agar and fresh liver was replicated 4 times temporally in the same room. Fifty puparia were randomly collected from each

PAGE 27

27 of the three containers and were individually placed into a 29.6 mL plastic cup with a tight fitting lid until eclosi on or mortality was confirmed. Surviving adults were counted, sexed and added to the colony. Results and Discussion Although li ver is the standard rearing substrate for large numbers of calliphorid larvae, it cannot be used in very small amounts (5 g) as it desiccates very quickly. Indeed, we found that 5 g of liver desiccated within 2 6 hours, resulting in 100% mortality (Table 2 1). In contrast, it was found that 5 g of liver agar remained moist for at least 12 hours, and survivorship on 5 g liver agar was comparable to that reported in other studies using larger pieces of liver (Goodbrod and Goff 1990, Gabr é et al. 2005), and to that on 50 g of liver or liver agar. Desiccation of liver and viability of liver agar have been observed repeatedly (personal obs.) over long term calliphorid rearing operations under widely varied conditions. In addition to >20 generations of C. megacep hala , I successfully reared Lucilia coeruleiviridis (Macquart), Phormia regina (Meigen), Cochliomyia macellaria (F.), Calliphora vicina Robineau Desvoidy and Chrysomya rufifacies (Macquart) adults from field collected larvae on this liver agar substrate. O ne of the advantages of the liver agar is that on the rare occasion that it was observed to be dehydrated, a few drops of water would rehydrate it for another 12+ hours, whereas previous observations indicated that liver would desiccate regardless o f wheth er we add water or not. Indeed, increasing humidity in the growth chambers or adding water to the liver has been seen to increase mold growth. The ability of the liver agar not to desiccate will be especially important at cooler temperatures, when developm ent time is slower and sample check frequency can be decreased.

PAGE 28

28 Table 2 1. Survivorship (percent 1 st instar larva to adult eclosion) of Chrysomya megacephala (F.) on fresh liver and liver agar substrates. Substrate # of Larvae Survivorship (%) SE 5 g fresh liver 10 0.0 5 g liver agar 10 83.3 1.2 50 g fresh liver 300 91.5 2.9 50 g liver agar 300 86.2 4.6

PAGE 29

29 CHAPTER 3 VOLUME OF LARVAE IS THE MOST IMPORTANT SINGLE PREDICTOR OF MASS TEMPERATURES IN THE FORENSICALLY IMPORTANT CALLIPHORID CHRYSOMYA MEGACEPHALA (DIPTERA: CALLIPHORIDAE) Introduction Chrysomya megacephala , the oriental latrine fly, is found throughout the tropical regions of the world, with recent expansion into more temperate areas (Greenberg 1988, Wells 1991, Martinez Sanchez et al. 2001, Tomberlin et al. 2001) . This forensically important blow fly is one of the first species of calliphorids to arrive at a corpse. In north central Florida, C. megacephala has displaced Lucilia coeruleiviridis as the dominant fly species found on a body in the spring, summer and fall (SVG, personal observation). Usually, a female calliphorid first deposits her eggs in the nose, eyes, ears or mouth of a corpse. After the eggs hatch, the feeding larvae complete three larval instars, and disperse to pupariate. To estimate a postmortem interval (PMI), forensic entomologists attempt to use the age structure of the insects found in association with the corpse and the temperature to which they were exposed. Acquisition of weather data from nearby airports is usually the source of temperature data used to reconstruct temperature and age relationships. Calliphorid larvae form aggregations known as maggot ma sses , which vary depending on factors such as maggot mass size and location of the carcass , and the number of ovipositing female flies . It is well known that these masses are capable of generating heat (Cianci and Sheldon 1990, Anderson 2000, Marchenko 200 1, Nabity et al. 2006, Slone and Gruner 2007, Heaton et al. 2014), but the exact mechanism for how the larvae accomplish this is not known. Mass temperatures have been reported as high

PAGE 30

30 as 50 °C (Gruner and Slone 2007), but larvae do not remain in these hot zones for prolonged periods of time. Larvae in large masses continually move from the hot center to the cooler edges of the mass (Anderson and VanLaerhoven 1996 ; SV G , personal observation). T here are possible evolutionary benefits to feeding in large maggot masses. While feeding, calliphorid larvae produce digestive enzymes that break down animal tissues. These enzymes may increase feeding efficiency , especially in large maggot masses (Goodbrod and Goff 1990, Greenberg and Kunich 2002, Ireland and Tur ner 2005, Heaton et al. 2014). More individuals in a mass would result in greater enzyme output and higher maggot mass temperatures (Rivers et al. 2011). The elevated thermal environment may protect the larvae from unexpected drops in temperature (Cragg 19 56, Campobasso et al. 2001, Huntington et al. 2007). Cianci and Sheldon (1990) also suggested that heat produced from large maggot masses may reduce predation by shortening larval development times. Deonier (1940) found that maggot mass temperatures on go at carcasses were that of the atmosphere , and Turner and Howard (1992) found that larval mass temperatures on rabbit carcasses could be 20 26 °C higher than ambient. On pig carcasses (Slone and Gruner 2007) , it was found that ambient temperat ures had a significant effect on small larval masses but not on larval masses greater than 20 cm 3 . Larvae may move in and out of masses in order to cool off when the mass temperatures could be lethal to them , but the exact mechanism of larval cooling has y et to be studied (Rivers et al. 2011) . In a study by Slone and Gruner ( 2007), maggot mass temperatures

PAGE 31

31 were occasionally observed to be cooler than ambient temperatures. This may have been the resul t of evaporative cooling (Toolson 1987, Prange 1995, Rivers et al. 2011). Several authors (Joy et al. 2006, Rivers et al. 2 010, Charabidze et al. 2011, Heaton et al. 2014) reported that elevated maggot mass temperatures of species other than C. megacephala were determined by larval stage (instar) and number of individuals in a mass, with second instars not able to produce as much heat as third instar larvae. They determined that there was a strong correlation between mass size and generation of elevated magg ot mass temperatures, with higher mass temperatures associated with larger mass size. Using pigs as animal models in field studies conducted in 2006, Slone and Gruner ( 2007) determined that larval mass volume (> 20 cm 3 ) was a predictor of internal tempera ture, and that age or larval instar had no effect on maggot mass temperatures after taking volume into account . However, the aggregations used in the analyses consisted mainly of combinations of instars (1 2, 1 2 3, and 2 3). The objective of this laborato ry study , therefore, was to determine if maggot mass volume is the principal predict or of maggot mass temperatures by testing 12 models that systematically partition the relative contributions of ambient temperature, maggot age and number , and mass volume. Methods and M aterials Insect C ulture Chrysomya megacephala adults were maintained at 28 °C and 50 65% RH with 14L:10D photoperiod. They were fed powdered milk and sugar (50:50) ad libitum . For acquisition of eggs, a pproximately 50 g of fresh cow liver for oviposition substrate was added to cages containing about 200 adult flies for a period of about four hours .

PAGE 32

32 Rubbermaid ® (Atlanta, Georgia) c ontainers of different sizes were used during testing. Each container had a tight fitting top to prevent larvae from escaping. A circular size piece (appropriate for each container size) was cut out of each container lid and gauze was hot glued over it to allow for air ventilation. From past experience of rearing larvae in a laboratory environment, it was discovered that their ammonia secretions can cause toxic shock or even be lethal to the larvae if containers are not ventilated well. For masses of 40 and 100 larvae, 1.4 L containers were used , 4.0 L containers were used for 250 and 600 larvae, and a 7.8 L container was used for 2 000 larva e . Each container was half filled with fine vermiculite (Therm O Rock East, Inc., New Eagle, PA) . To continually monit or the temperature inside each container, a Hobo TCMC6 HA temperature probe (Onset Computer Corporation, Bourne, MA) was placed on the liver in the same approximate location as where the larvae were placed. Each temperature probe was connected to a Hobo H0 8 data logger set to record temperatures in 15 min intervals. For each replication, five Rubberma id containers with 40, 100, 250, 600 and 2000 larvae were randomly placed into two Florida Reach In chambers (Walker et al. 1993) set to a constant temperature of 20 °C, and five containers with 40, 100, 250, 600 and 2000 larvae were randomly placed in two chambers set to a constant temperature of 30 °C. Thus, for each replication, the temperatures, volumes and instars of larvae in every container (10 per test) were being monitored for the duratio n of their development until larvae reached the post feeding stage. This test was repeated five times for a total of six replicates.

PAGE 33

33 To determine maggot mass temperatures, external mass temperatures were meas ured with a Raytek Tayner ST infrared thermometer (Raytek Corporation, Santa Cruz, CA), and internal mass temperatures were measured with a Taylor ® 9841 digital pocket thermometer (Forestry Suppliers, Jackson, MS). Larval mass temperatures , combinations of instars present, and volumes of the developing larvae in all 10 containers were measured approximately every six to eight hours. Effect of Number of L arvae in a M ass To determine effect of number of larvae in a mass, newly hatched first instar larvae were counted in to five masses containing 40, 100, 250, 600, and 2000 larvae. The larvae were inoculated onto fresh cow liver in 10 containers. L iver was added continually through the experiments so there was always more than adequate food for the larvae to con sume . Effect of L arval I nstar in a M ass For both temperat ure treatments being tested (20 and 30 ° C) , larval mass temperatures were measured for each larval mass size when the re were approximately 100% first instars, 50/50 first/second instars, 100% second instars, 50/50 second/third instars, and 100% third instars. Effect of L arval V olume in a M ass To determine effect of volu me, volume of each developing larval mass (L x H x W cm) was measured with a ruler at each sample check. The Taylor ® thermometer was inserted into different regions of each mass and the depth at which the shaft penetrated was then measured. A cold temperature probe inserted into a maggot mass would result in immediate dispersion of the larvae in a mass (SVG, personal observation), so t he shaft of the thermometer was hand warmed before measurement. The probe was

PAGE 34

34 placed in each mass for approximately 30 seconds to measure the maggot mass temperature. The depth, length and width of each mass were recorded on a s preadsheet at each check tim e. These measurements, along with photographs of the masses, were used to estimate the volume of the mass. Each experiment ended when the larvae reached the post feeding stage. Although Hobo temperature probes were inserted into the liver where the larvae were initially placed, the larvae moved away from the probes. Attempts were made to relocate the probes to the center of the larval masses at each sample check time, but in almost every instance, the larvae moved away from the probes. T he maggot mass temperatures of first instar larvae in aggregations of 250 or fewer were difficult to obtain with the Taylor thermometer . For these smaller maggot masses, the external temperatures taken with the infrared thermometer were used in the data analyses . Environmental Chamber Temperature Profile Prior to all testing, each rearing chamber was monitored for spatial and temporal tempe rature variability with data loggers connected to Hobo temperature pr obes. Eleven probes were placed on/around the center shelf of each unit and temperatures were recorded at 15 min intervals for 48 h to determine if there were any temperature variations within each unit. Temperature probes were calibrated by the manufactur er, and before use, we verified that all probes were within 0.01 °C of each other when placed in the same environment. Statistical A nalyses The parameters of interest for the modeling were chosen to represent the competing theories of maggot mass thermoge neration. They included number of larvae,

PAGE 35

35 instar, volume, and ambient temperature . Box and whisker plots were generated to represent the overall patterns in the data dis t ribut i ons. For volume analysis, developing masses were measured at each check time an d categorized into the following mass sizes: 2, 2.1 7, 7.1 20, 20.1 50, 50.1 150, 150.1 400, 400.1 1100 and 1100.1 5600 cm 3 , irrespective of age. Maggot mass temperatures, number of maggots, and volume were log transformed to normalize the data . Preliminary data exploration showed a curved pattern in volume and elapsed time, so a quadratic term was added for these variables. Twelve different multiple regression candidate models were selected which included a container identification number ( CID) for proper partitioning of repeated measurement error. The number and instar of the larvae in a mass both are correlated with the volume of the mass larval instars, and cannot be included together as predictors in a traditional general linear mixed mo del (GLMM), but other combinations of parameters as suggested in the literature were included in the candidates. These models were a nalyzed with R using package lme4 (Bates et al. 2014) , then ranked and scored with the Akaike information criterion ( AIC ) and Bayesian informat ion criterion (BIC) (Burnham and Anderson 2001, 2004), which are model selection techniques that combine model fit and complexity. The model with the lowest score is considered to be the best. Graphs were generated with R Statistical Software version 3.1.0 (R Core Team 201 4 ) using ggplot2 (Wickham 2009 ) ( R codes in Appendix A ) .

PAGE 36

36 Results Environmental Chamber Temperatures Actual temperatures in all units were higher than set temperatures. The set/actual temperature averages were 20/21.2 and 30/30.8 °C. The actual average chamber temperatures were used for all analyses. Number of Larvae in a M ass Larvae formed aggregations within minutes of being placed onto the liver. In all cases, as the number of larvae increased in a mass, the median temperature of the masses also increased. Generally, the variability was greater in the larger masses (Figure 3 1). At the ambient background temperature of 21.2 °C, median temperature s of all mass sizes were above ambient. Masses containing 40, 100 and 600 larvae had temperatures lower than ambient in 19% of the measurements. At 30.8 °C, all mass sizes had temperatures lower than ambient , and the masses containing 40 and 100 larvae had median temperatures lower than ambient. Larval Instar s in a Mass At 21.2 °C, median mass temperatures were generally higher in later instars but there was only a 5 °C difference between first instars and third instars (Figure 3 2). Still, the median tem peratures of all instar classes were above ambient, and the minimum temperature s of all masses were at or slightly lower than ambient. T emperatures in the low end of the whisker intervals for all masses were at ambient or lower than ambient. As was the cas e with smaller masses, in the lower instar classes the median temperatures of all instar classes were above ambient.

PAGE 37

37 At the higher ambient temperature of 30.8 °C, the median temperature of third instars, and combination of second and third instars, was gr eater than ambient, but the earlier instar classes had median temperatures less than ambient. This also is similar to the trend seen with number of larvae in a mass. The minimum temperature for all of the instar classes was very similar, but the ranges of mass temperatures were higher in the masses with older instars. Maggot Mass Volume At both ambient temperatures, larval masses with greater volumes had higher median temperatures, except masses containing about 5600 cm 3 at 30.8 °C (Fig ure 3 3) . The mass containing about 5600 cm 3 of larvae had a median temperature slightly less than that of the mass containing about 1100 cm 3 of larvae, but still had a median temperature that was higher than all other mass sizes. At both ambient temperatures, the re wa s a trend for the variability in mass temperatures to increase as mass volume increased. Predictive M odels of Maggot Mass Temperature Maggot mass temperatures were best predicted with models using mass volume as a parameter (Table 3 1, models 4, 9 and 11). Models predicting mass temperature solely as a function of number of maggots (Model 1), or instar (Model 2), or treatment temperature (Model 3) did not score well with AIC or BIC. Addition of elapsed time to model 11 resulted in a lower AIC score than Model 9 (indicating a better model), but it added 12 parameters to the model, thus raising the BIC rank (indicating a poorer model).

PAGE 38

38 Discussion AIC and BIC are model selection techniques; each consists of a goodness of fit term and a penalty to control for overfitting (Burnham and Anderson 2001, 200 4) . BIC tends to favor models with fewer parameters than AIC (Burnham and Anderson 2004) . The AIC scores indicated that model 11 was better than model 9 but included the addition of 12 parameters to the model. Model 9 is the most parsimonious based upon the combination of AIC score and BIC rank. T he results of this lab oratory study indicate that although instar ( age ) , treatment temperature (21.2 and 30.8 ºC) and number of maggots significant ly affected maggot mass temperature , larval volume by itsel f was the best single predictor of maggot mass temperatures . Volume was the principal factor affecting mass temperature, but the model was improved significantly by addition of temperature treatment. This is consistent with the results of the field study b y Slone and Gruner ( 2007) in which it was determined that volume was the principal predictor of maggot mass temperatures (in other species of calliphorids) in field deposited pigs, rather than instar, larval density, ambient temperature, or number of maggo ts. As observed in the field study , at higher ambient temperatures, the larval m ass temperatures were below ambient at times (Slone and Gruner 2007). At the ambient laboratory temperature of 30.8 °C, median temperatures of mass sizes smaller than about 50 cm 3 were less than ambient. Both small and large masses may be using evaporative cooling to regulate their temperatures, but the exact mechanism for how cooling is achieved has yet to be studied in depth (Toolson 1987, Rivers et a l . 2011). In contrast, a t both ambient temperatures , larval masses > 150 cm 3 sometimes attained

PAGE 39

39 temperatures over 40.0 °C . At the highest rearing temperatures, the largest mass had temperatures of approximately 46 °C in the upper end of the whisker interval . Heaton et al. (2014) , Rivers et al. (2010), and Charabadze et al. (2011) determined that number of maggots and instar were the principal predictors of maggot mass temperatures, but vo lume was not measured in these studies. The results of this study are not so different becaus e the number of maggots and instars by their age, size, or weight are highly correlated with volume. A s previously discussed (Slone and Gruner 2011, Rivers and Brogan 2011), keeping larval numbers constant over time will result in greater mass temperatures because the larvae are simply getting larger, therefore increasing the volume. Larval mass volumes are not that difficult to estimate, using the method developed here . T his is the first study to partition the relative contributions of volume, larval insta r (age) and number of maggots to mass temperature prediction. As mentioned by Heaton et al. (2014), the volume or size of larval masses is usually not mentioned in reports evaluating larval development for estimations of a PMI (Benecke 1998, Introna et al. 1998) . Some published calliphorid development studies (Greenberg 1991, Anderson 2000, Byrd and Allen 2001) were conducted us ing masses large enough to generate heat above that of the set environmental chamber temperatures. VanLaerhoven (2008) stated that the effect of elevated mass temperatures are taken in account in such studies, and therefore is not an issue. But this may be misleadin g because larval mass temperatures will affect development times and the temperature at which larvae were being tested may not have been the actual temperatures they were experiencing for the duration of the test. The best method to

PAGE 40

40 determine effect of tem perature on development is therefore to remove mass heating as a variable by using small numbers of larvae (20 or fewer).

PAGE 41

41 Figure 3 1. Larval mass temperatures as a function of number of larvae in a mass at two different treatment temperatures (21.2 and 30.8 °C). The boxes represent the interquartile ranges of the data; the black horizontal bars represent the median temperatures. The vertical lines represent 1.5 × the interquartile range, which approximate s a 95% confidence interva l of the data and the black diamonds are outliers. At 30 °C, the median values of the masses containing 40 and 100 larvae were less than ambient , but all larval masses generated temperatures greater than ambient.

PAGE 42

42 Figure 3 2. Larval mass temperature s as a function of larval instar at two different ambient temperatures (21.2 and 30.8 °C). The boxes represent the interquartile ranges of the data; the black horizontal bars represent the median temperatures. The vertical lines represent 1.5 × the interqu artile range, which approximates a 95% confidence interval of the data and the black diamonds are outliers. The numbers on the x axis correspond to instars and approximate combinations thereof: 1 represents 100% first instars, 1.5 represents 50/50 first an d second instars, 2 represents 100% second instars, 2.5 represents 50/50 second and third instars, and 3 represents 100% third instar larvae.

PAGE 43

43 Figure 3 3 . Larval mass temperature as a function of volume. L arval masses with greater volumes had higher median temperatures . The boxes represent the interquartile ranges of the data; the black horizontal bars represent the median temperatures. The vertical lines represent 1.5 × the interquartile range, which approximates a 95% confidence interval of the data and the black diamonds are outliers.

PAGE 44

44 Table 3 1. Twelve candidate models and their AIC and BIC scores. The model with the lowest score is considered to be the best. Model Parameters # of Parameter s AIC BIC 1 log (MMT) = I + CID + log (# Maggots) 3 221.93 207.04 2 log (MMT) = I + CID + Instar 3 374.55 359.65 3 log (MMT) = I + CID + TT 3 241.48 226.58 4 log (MMT) = I + CID + (log (V) + log (V)²) 4 486.79 468.17 5 log (MMT) = I + CID + Instar * log (# Maggots) 5 436.31 413.97 6 log (MMT) = I + CID + TT * Instar 5 416.46 394.12 7 log (MMT) = I + CID + TT * log (# Maggots) 5 313.34 291.00 8 log (MMT) = I + CID + TT * (ET * ET²) 7 415.28 385.49 9 log (MMT) = I + CID + TT * (log (V) + log (V)²) 7 606.77 576.98 10 log (MMT) = I + CID + TT * Instar * log (# Maggots) 9 549.40 512.16 11 log (MMT) = I + CID + TT * (log (V) + log (V)²) * (ET + ET²) 19 622.99 548.52 12 log (MMT) = I + CID + TT * Instar * log (# Maggots) * (ET + ET²) 25 576.73 479.92 MMT = Maggot Mass Temperature V = Volume CID = Container Identification Number TT = Treatment Temperature I = Intercept ET = Elapsed Time * denotes interaction, + denotes no interaction

PAGE 45

45 CHAPTER 4 DEVELOPMENT OF THE ORIENTAL LATRINE FLY, CHRYSOMYA MEGACEPHALA (FABRICIUS) (DIPTERA: CALLIPHORIDAE) AT FIVE CONSTANT TEMPERATURES Introduction Chrysomya megacephala (Fabricius), the oriental latrine fly, is found throughout much of the tropics and subtrop ics worldwide. The geographical range of this blow fly was originally the Australasian and Pacific regions but this species has expanded its range to the Palearctic, Nearctic and Neotropical regions (Baumgartner 1988, Goff 1991, Wells and Kurahashi 1994) . Within North America, its range now includes California (Greenberg 1988) , Texas (Wells 1991) , Florida (Byrd and Butler 1997, Gruner et al. 2007) , Louisiana (Pharr 2009) and Georgia (T omberlin et al. 2001) . Chrysomya megacephala is a filth fly that can cause secondary myiasis in humans and animals (Zumpt 1965, Sukontason et al. 2005) . It also can transmit enteric pathogens and parasites, and is the dominant insect vector of helminth eggs in Malaysia (Sulaiman et al. 1988, 1989) . This blow fly is also an important tool in forensic entomology as it is one of the first species of calliphorid flies to arrive at a corpse. In north central Florida, Lucilia coeruleiviridis (Macquart) was formerly the first calliphorid to arrive at a corpse and was the dominant calliphorid found in maggot masses (Gruner et al. 2007) , but Chrysomya megacephala has replaced L. coeruleiviridis in the spring, su mmer and fall months (personal observation). For regions in which C. megacephala is present, knowledge of temperature dependent development is a key factor in estimating the ages of specimens collected from a corpse . F rom this age, estimation a post mortem interval (PMI) can be inferred.

PAGE 46

46 Although there are studies of development of C. megacephala (Bohart and Gressitt 1951, Khole 1979, Subramanian and M ohan 1980, O'Flynn 1983, Nishida 1984, Goodbrod and Goff 1990, Wells and Kurahashi 1994, Gabre et al. 2005) , data from the early studies (pre 1980s) were not meant to be extrapolated for use in determination of a PMI in forensic entomology. The work of Wells and Kurahashi (1994) and Gabré et al. (2005) present useful information of the life cycle of C. megacephala , but development at only one temperature was tested. Only Nishida (1984) tested development as a function of temperature. Other aspects of C. megacephala are similarl y poorly documented, so laboratory based observations were made on reproductive biology. There are development rate studies for other forensically important calliphorids such as Phormia regina ( Meigen ) (Kamal 1958, Nishida 1984, Greenberg 1991, Introna et al. 1991, Anderson 2000, Byrd and Allen 2001) , Calliphora vicina ( Robineau Desvoidy) (Kamal 1958, Greenberg 1991, Introna et al. 1991, Anderson 2000, Ames and Turner 2003, Do novan et al. 2006) , Chrysomya rufifacies (Macquart) (Greenberg 1991, Byrd and Butler 1997) , Cochliomyia macellaria (Fabricius) (Greenberg 1991, Byrd and Butler 1997 ) , and Lucilia sericata (Meigen) (Kamal 1958, Greenberg 1991, Anderson 2000, Grassberger and Reiter 2001) but there is no standardized method of obtaining development rates or times. Variables such as ambient temperatures, diet and substrate, humidity, photoperiod, constant vs. fluctuating temperatures, resolution various studies. A major problem with m ost of the published development rate studies is that they lack replication. Additionally, some studies focused on development of the first

PAGE 47

47 10% or the each life stage ( Byrd and Butler 1997, Grassberger and Reiter 2001, Greenberg 1991) . The number of eggs or larvae used to start a test is a very important factor when determining development rates for use in age estimations. It is well documented that calliphorid larvae usually form dense aggregations and that these agg regations generate heat (Deonier 1940, Early and Goff 1986) . Larval masses with as few as 4 0 100 larvae can generate enough heat to raise the temperatures above ambient temperatures (Gallagher et al. 2010, Gruner C hapter 3, Heaton et al. 2014) . One way to measure more precise temperature related development times is to use small numbers of larvae for each temperature being studied. Thus, to determine development times of C. megacephala without the effect of larva l heat generation, masses of only 10 larvae were evaluated, thereby excluding this confounding factor. When estimating a PMI, it may be necessary to account for the elevated mass temperatures. Slone and Gruner (2007) reported that mass temp eratures depend largely on the volume of the larval masses. The masses in their studies contained multiple species of calliphorid larvae that included: L . coeruleiviridis , C . rufifacies , C . macellaria , P . regina and C. megacephala . It is possible that each species has an optimal development temperature, but the communal mean maximum maggot mass temperatures obtained during the field study were between 38.3 and 44.6 °C. There have been few published calliphorid development studies with testing conducted at o r higher than 35 °C. Byrd and Allen (2001) reported that development of P. regina occurred at 40 °C (and was faster than development at 35 °C) constant temperature,

PAGE 48

48 but there was no adult emergence after the puparial stage; thus, 40 °C approaches the upper lethal limit for that species. The objective of this lab study was to calculate the accumulated development time and population transition points for C. megacephala from oviposition for each life stage from eclosion to adult emergence at eight constant temperatures: 5, 10, 15 , 20 , 25 , 30 , 35, and 40 ° C . As larvae in very small masses are unable to thermoregulate (Slone and Gruner 2007), the effect of the maggot mass temperatures on development time was eliminated by having only 10 larvae per cup. Ambient and larval mass temperatures were collected from field studies conducted from 2001 2004 using pigs , Sus scrofa (L) , as human models to obtain estimates of maggot mass elevated temperatures under field conditions. The objective of this research was to dete rmine if the combination of C . megacephala transition times with the timing of elevated mass temperatures determined in field studies could be used for a more precise estimation of a postmortem interval (PMI). M aterials and Methods The C. megacephala colon y was started with adult flies collected from a death scene in Jacksonville, Florida in July, 2006. Approximately 200 wild C. megacephala adults from various sites in north central Florida were added to the colony each summer until research started in May, 2010. Approximately 2,000 3,000 adult flies were maintained in screen cages (28 x 28 x 51 cm) in a rearing room at 28 ± 2 °C with 50 65% RH and a photoperiod of 14L:10D , with approximately 200 flies kept in each cage. Adult flies were fed a 50:50 mixture of table sugar:powdered milk, with fresh water supplied ad libitum .

PAGE 49

49 Environmental Chamber Temperature Profile Temperature studies were conducted in Florida Reach In chambers (Walker et al. 1993) . Each chamber was monitored for spati al and temporal temperature variability with Hobo ® H8 (Onset Computer Corporation, Bourne, MA) data loggers connected to Hobo ® TCMC6 HA temperature probes. Eleven probes were placed on/around the center shelf of each unit and temperatures were recorded at 15 min intervals for 48 h to determine if there were any temperature variations within each unit. Temperature probes were calibrated by the manufacturer, and before use, it was verified that all probes were within 0.01 °C of each other when placed in the same environment. D evelopment times were tested in two trials, each of which included two chambers for each temperature tested. Each chamber had one vented 27 L Sterilite ® covered container with 6 cups that contained 10 larvae per cup. To minimize the effect of seasonality, th e order in which temperatures were tested was randomized (5 °C: November, 2010, March, 2011; 10 °C: November, 2010, April 2011; 15 °C: April 2011, November 2011; 20 °C: March 2011, September, 2011; 25 °C: February, 2011, September, 2011; 30 °C: November, 2010, October 2011; 35 °C: February, 2011, September, 2011, and 40 °C: November 2010, September, 2011). During each trial, three probes were placed on the front, left side and back of the cover of each large container, and one probe was placed in the cup nearest the center of the 27 L container. T he environmental chambers were set to constant temperatures of 5, 10, 15, 20, 25, 30, 35, and 40 °C. For the 5, 10 and 15 °C replicates, the temperature in the main r earing room was gradually lowered from 28 °C to 24 °C for at least 4 d prior to egg collection to acclimate the flies to slightly cooler temperatures.

PAGE 50

50 Egg C ollection Cold Stora ge, Newberry, Florida) into the cages of adult C. megacephala flies. Liver was checked every 15 min for egg deposition. Once egg deposition began, the time was noted as T 0 and egg clusters were collected after 45 min. Thus, eggs collected were one hour old or less. Eggs were separated into two groups and placed into the center cup in each of the two chambers set to the trial temperature. Laboratory Development Time Tests After egg hatch, the two groups of first instar larvae were randomized by mixing with a fine paintbrush (Royal SG 250 size 4, Royal & Langnickel Manufacturing, Munster, IN) . With the brush, 10 first instar larvae were collected and placed onto 5 g of fresh liver agar (Grun er and Slone 2014) inside a 100 mm x 15 mm polystyrene Petri dish (Fisherbrand ® , Fisher Scientific, Pittsburgh, PA) ( Figure 4 1 A). Fresh liver agar was added ad libitum . Petri dish tops were modified by burning a 3 cm diameter hole over which fine mesh was hot glued to allow for air circulation ( Figure 4 1B) . To keep humidity high, a small piece of wet paper toweling was set on a smaller 2.1 cm polystyrene Petri dish (Fisherbrand ® ) contained within the larger dish ( Figure 4 1B). A tight fitting elastic b and was wrapped around each large petri dish to prevent larvae from escaping ( Figure 4 1B). Each Petri dish combination was placed into a 0.47 L plastic cup (Dart ® , Mason, MI) that was half filled with fine vermiculite ( Figure 4 1C). Each plastic cup had a tight fitting lid with an approximately 5 cm hole covered with fine mesh to allow for air circulation ( Figure 4 1C).

PAGE 51

51 Six cups were placed into a lidded 27 L plastic container (Sterilite ® , Townsend, MA) in which two 40 x 5 cm holes (one on each side of th e container) had been cut and covered with fine mesh to allow for air circulation. Approximately 50 water saturated paper towels were placed at each end inside each container in order to maintain high humidity. Each 27 L container was then placed on the ce nter shelf of a Florida Reach In chamber. Two large containers (Rubbermaid ® 7.8 L) were filled with water and placed on the bottom shelf of each chamber to maintain approximately 60% RH at all times during the testing. Humidity was measured with an LCD dig ital hygrometer (Cole Palmer Instrument Co., Chicago, IL). Sample s were checked at a maximum of 12 h intervals, but as frequent as hourly around expected stage transitions. At each sample check, the number of larvae at each life stage was recorded . Larvae were individually counted and identified under a dissecting microscope (Leica StereoZoom ® 7). Transitional i nstars were considered to be the previous instar (i.e., a 1 2 transitional instar was counted as a first instar larva). As larvae developed to the wandering stage, the Petri dish tops were removed and the dish with wet paper toweling was placed on top of the vermiculite to allow the larvae to move into the vermiculite to pupariate. For examination and counting of postfeeding larvae and pupae, the ver miculite was gently sifted into a pan, and any puparia found were placed onto a separate small Petri dish inside the plastic cup. After adults emerged, they were sexed, and then added to the existing colony. Occasionally, all 10 first instar larvae could n ot be located, so only the development stages of the visible larvae were recorded.

PAGE 52

52 C hrysomya megacephala larvae were observed to be negatively phototropic . Larvae clustered together under the liver agar most of the time. The minute first instar larvae woul d cluster together almost immediately after being inoculated onto the liver agar, even if collected and placed one at a time. When fresh agar was added, larvae usually moved onto it, and old agar was removed. Field Data Collection From 2001 to 2004, pig ( Sus scrofa L.) carcasses were placed in a wooded area near Earleton, FL, as part of a calliphorid fly succession and thermal biology research project. Ambient temperatures were collected daily with a Taylor 9841 digital thermometer (Forestry Suppliers, Jac kson, MS). Once mass aggregations formed, internal mass temperatures (from 5 to 50 locations per mass, depending on mass size) were collected between the hours of 1400 and 1700 hours with the Taylor thermometer. External mass temperatures were collected wi th a Raytek infrared thermometer (Forestry Suppliers, Jackson, MS). During warm months, meas urements were taken daily, but in cooler months, mass temperatures were measured every second or third day. Detailed methodology is found in Slone and Gruner (2007) . From the unpublished data , daily ambient temperatures taken at the beginning of each sample check were obtained. For each pig (N = 38), mean maximum mass temperatures, peak maggot mass temperatures, days until the maggot mass temperatures were elevated a bove ambient, duration of peak mass temperatures , and percent increase of mass temperature above ambient were determined . Statistical Analyses Development time was transformed with square root to normalize the data and then analyzed with R Statistical Software version 2.15.2 (R Core Team 2012) using

PAGE 53

53 logistic regression to estimate the transition times for each stage at each tested temperature. Parameters were: elapsed time, transition number, temperature, environmental chamber number , cup number and tri al number. For each transition time, the 10 th , 50 th (median) and 90 th percentile (± SE) were modeled (Appendix B) , then all results were back transformed for display purposes (R code in Appendix C ) . The mean transition time (± SD) was calculated from the r aw laboratory data . The temperatures recorded from the data loggers within the chambers were averaged for each treatment. The actual temperatures collected by the data loggers were used for analysis. Percent survivorship was calculated from the raw data a s the mean number of surviving adults divided by the initial number of first instar larvae per t emperature per trial. Ribbon graphs to compare development results to other studies were generated using the calculated percentiles with R using ggplot2 (Wickham 2009) . Field trial data were separated into five temperature groups ( 14.1 17.5, 17.6 22.5, 22.6 27.5, 27.6 32.5, and 32.6 34.5 ) based upon the ambient temperature at the time the pigs were deposited at the site. Mean maximum larval mass temperatures, days until larval mass temperatures were elevated, and mean daily peak larval mass temperatures were determined using the raw data. To determine percent increase ove r ambient temperatures, the mean of each of the five ambient temperature groupings ( 15.8, 20.0, 25.0, 30.0, and 33.5 ) and the means of the maximum mass temperatures were used . Results Environmental Chambers A ctual temperat ures in all rearing units were slightly higher than set temperatures . The set/actual temperature average s were 15/16.0, 20/21.2, 25/25.8,

PAGE 54

54 30/30.8, 35/35.6 °C. The actual difference in set/actual chamber temperatures varied diurnally, with a 1°C or less dif ference at night and a 1.5 °C or less difference during the day, and was consistent among chambers . The temperature variation among the chamber probes, including the probes in the rearing cups, was <1 °C . The actual chamber temperatures, based on the means of the probes inside and on the cover of the 27 L container, were used for all development time analyses. Laboratory Development without Effect of Maggot Mass A total of 1200 larvae in 120 cups were sampled . Eggs did not survive at 5, 10 and 40 °C; thus, development times were not obtained at these temperatures. Cumulative times to each life stage transition at five constant temperatures are presented in Tables 4 1 4 5. Temperature variations within and between environmental chambers are generally not t ested (Nabity et al. 2006), but h ad the temperature inside each environmental chamber not been measured and accounted for temperature error prior to testing , any predictions made from the development data would miscalculate development time by as much as 5 2 hours, assuming a chamber error of 1 °C (Table 4 6). Development was curvilinear. As the ambient temperature increased, larval development times decreased . Additionally, as the ambient temperature increased, the gap between the 10 th and 90 th percentiles narrowed. At the lowest temperature of 16.0 ºC, the mean development time from egg to adult was twice that of the mean development at 21.2 ºC. Mean total development time at the highest temperature of 35.6 ºC was only 12.1 h faster than mean t otal development at 30.8 ºC. Survivorship at each temperature was 14% (16 °C), 83% (21.2 °C), 85% (25.8 °C), 67% (30.8 °C) and 55% (35.6 °C). The larvae developed fastest at 35.6 °C. Based

PAGE 55

55 on survivorship, the optimal growth temperature for C. megacep hala may be about 25.0 °C ambient temperature. Field Trials with Effect of Maggot Mass Temperatures Larval mass temperatures were not elevated above ambient on carcasses in the field until about two to six days postmortem when ambient temperatures were b elow 17. 5 °C, but were present with elevated temperatures after only one or two days at temperatures higher than 27.6 °C ambient . At temperatures greater than 32.6 °C, larval masses had temperatures elevated above ambient on the second day postmortem . The duration of elevated mass temperature was shorter at the higher ambient temperature s because the larvae developed and dispersed more rapidly. T he mean maximum mass temperature that the masses attained was also higher with increasing ambient temperatures , b ut the peak temperature was at the middle temperature range of 2 2.6 27.5 °C (Table 4 7 ) . Discussion It was determined that adults lived as long as three months in colony at 28 °C. Adult females were observed to begin oviposition 20 days after emergence, with the majority of oviposition occurring on the 21st day post emergence. Adults were not active each day until the mid afternoon, which has also been observed in the fiel d (SVG personal observation). After establishing the laboratory colony, oviposition substrate was not placed into the cages until late in the afternoon because females would not oviposit on it until 15:00 regardless of when the substrate had been placed. Esser (1991) found that the average longevity of three generations of C. megacephala was 47 to 55 days, and that maximum adult longevity could be as many as 98 days at 29 °C. This is similar to my personal observations that adults in colony

PAGE 56

56 lived as long a s three months at 28 °C; however, these results differ from that Gabré et al. (2005), who found mean adult longevity to be only 25 days. Additionally, my observations of the adults in colony indicated that females were not gravid until 19 days after emerge nce, which is similar to what Esser (1991) observed. Wells and Kurahashi (1994) found that C. megacephala eggs hatched after 12 18 h and that oviposition to adulthood took 198 234 h at 27°C. This is similar to what w as measured in this study at 25.8 ° C ( Figure 4 2). The results of this research seem C. megacephala at 28°C, Bohart and Gressitt (1951), who tested at 29.4 °C, and Khloe (1979) who tested at 27 °C. Subramanian and Mohan (1980) tested at 25 .6 °C and their results indicate a faster development time than what was measured during this test at 25.8 °C, indicating possible larval heating in their study. Nishida (1984) observed oviposition to pupariation periods of 12 d at 24 °C, 11 d at 30.8 °C and 8 d at 35.6 °C ; drastically different than my results that showed oviposition to pupariation time s of 5 d at 25.8 °C, 4.1 d at 30.8 °C, and 4 d at 35.6 °C ( Figure 4 2 ). One can only speculate that the specimens tested in the Nishida study might have exp erienced suboptimal conditions . Chrysomya rufifac i es , the hairy maggot blow fly, is closely related to C. megacephala and is also a warm weather fly. Development of C. megace phala was compared to published data on C. rufifac i es and found that their development times were similar ( Figure 4 3 ). But, Nelson et al. (2009) found significant differences in development of sister species C. megacephala and C . saffranea (Bigot); thus, it should not be assumed that closely related species have sim ilar development times.

PAGE 57

57 A lso compared were C. megacephala development times to that of Lucilia sericata ( Figure 4 4 ), a calliphorid fly that is found in both tropical and temperate regions of the world. The data comparisons for most of the published data indicate that development times for both species are very similar. Importance of Ambient and Maggot Mass Temperatures for PMI Estimations The most important information that a forensic entomologist should collect at a death scene are ambient temperatures, larval mass volumes and temperatures and insect specimens. It is also necessary to determine what the amb ient temperatures were at the scene before the body was found; this is usually accomplished by accessing databases from nearby airport weather stations. Knowledge of the timing of elevated mass temperatures may be useful when estimating a PMI. At lower am bient temperatures, maggot mass temperatures can be considerably elevated, whereas at higher ambient temperatures, maggot mass temperatures are less elevated (Table 4 7). For example, during days when the ambient temperatures are < 17.5 °C, the mean mass t emperature can be more than twice that of ambient. At hotter ambient temperatures, maggot mass temperatures remain more similar to ambient temperatures; thus, any adjustment in maggot development times and subsequent assessment of PMI seems to be more crit ical at low ambient temperatures than at high ambient temperatures. As illustrated in Table 4 6, variations in environmental chambers can lead to errors in estimations of PMI. Environmental chamber temperature profiles are rarely determined before testin g (Nabity et al. 2006) ; thus, use of development data sets in which chamber temperature profiles were not tested should be used with caution.

PAGE 58

58 The importance of obtaining lab data without the effect of maggot mass temperatures is important especially at higher and lower ambient temperatures. Slone and Gruner (2007) reported that fly activity during the field trials was observed to vary in proportion to ambient temperature. When the ambient temperature was cool (< 2 0 °C) the number of flies ovipositing wa s very small, and at higher ambient temperatures, especially > 2 5 °C, the number of flies ovipositing was dramatically greater. The relative lack of fly activity in cooler ambient conditions explains, in part, why the mass temperatures were not elevated un til the fourth day postmortem. The peak maggot mass temperatures were not as high as the peak temperatures found at higher ambient temperatures because the larval mass volumes were smaller, on average, than those in hotter conditions. The peak mass tempera tures were not found on days with the highest ambient temperatures. Charabidze et al. (2011) determined that maximum heat emissions occurred when ambient temperatures were betw een 22 to 25 °C, which is similar to what w as found during the field trials ( Table 4 7 ). Knowledge of the timing of elevated mass temperatures may be useful when estimating a PMI. At lower ambient temperatures, maggot mass temperatures can be considera bly elevated, whereas at higher ambient temperatures, maggot mass temperatures are less elevated (Table 4 7 ). For example, during days when the ambient temperatures are < 17.5 °C, the mean mass temperature can be more than twice that of ambient. At warmer ambient temperatures, maggot mass temperatures remain more similar to ambient temperatures; thus, any adjustment in maggot development times and subsequent assessment of PMI seems to be more critical at low ambient temperatures than at high ambient tempera tures.

PAGE 59

59 T he objective of this study was to obtain accurate development times for C. megacephala to be used in estimations of a PMI. To date, this research provides the most detailed and comprehensive d ata of C. megacephala. To apply results from this laboratory study to field collected data, the temperature associated with maggot masses must be considered before an accurate development time can be calculated (Slone and Gruner 2007). The field study results suggest that at low ambient temperatures, the adjustment for mass based temperature elevation would be greater relative to high ambient temperature. Clearly, use of only ambient temperature in a model would result in overestimation of a PMI. F urther research is needed to determine how to accurately ac count for the elevated mass temperatures in development models, such as degree day models

PAGE 60

60 Figure 4 1. Cup arrangement (Photo courtesy of author). A) Uncovered Petri dish containing wet paper t oweling and 5 g of liver agar. B ) Covered Petri dish (sea led with an elastic band) containing 5 g of liver agar and wet paper toweling. C) Covered plastic

PAGE 61

61 Figure 4 2. Ribbon graph depicting development stages and transition times of C. megacephala compared to that o f other published development studies of C. megacephala . Colored circles on the horizontal ribbons correspond to time of stage determined by other author(s). The black dots represent the 50th percentile data from this study.

PAGE 62

62 Figure 4 3. Ribbon graph depicting development times of C. megacephala compared to development from published studies of C. rufifacies . Colored circles on the horizontal ribbons correspond to time of stage determined by other author(s). The black dots repr esent the 50th percentile transition time of C. megacephala from this study.

PAGE 63

63 Figure 4 4. Ribbon graph depicting our development times of C. megacephala compared to development from published studies of L. sericata . Colored circles on the horizontal ribbons correspond to time of stage determined by other author(s). The black dots represent the 50th percentile transition time of C. megacephala from this study.

PAGE 64

64 Table 4 1. Cumulative times from oviposition to each life stage transition for Chrysomy a megacephala at the 10th, 50th (median) and 90th percentiles (hours) and mean transition times (hours ± SD) at 16.0 °C constant temperature. Percentiles were back transformed from square root. Quantile values were predicted from modeling results, and mean s were calculated from laboratory data. Stage 10th 50th 90th Mean ± SD Egg to 1st instar 43.7 47.5 51.4 45.0 ± 6.3 1st instar to 2 nd 121.0 128.7 136.8 123.6 ± 3.3 2nd instar to 3 rd 201.0 210.4 220.1 203.1 ± 12.0 3rd instar to wandering 329.2 349.1 369.6 344.9 ± 24.5 Wandering to puparium 380.7 400.6 421.0 405.1 ± 26.8 Puparium to adult 731.1 768.9 807.6 719.2 ± 36.8 Table 4 2. Cumulative times from oviposition to each life stage transition for Chrysomya megacephala at the 10 th , 50 th (median) and 90 th percentiles (hours) and mean transition times (hours ± SD) at 21.2 °C constant temperature. Percentile values were back transformed from square root. Quantile values were predicted from modeling results, and means were calculated from laboratory data. Stage 10th 50th 90th Mean ± SD Egg to 1st instar 21.4 24.0 26.8 23.2 ± 1.4 1st instar to 2 nd 56.2 61.5 67.1 58.3 ± 2.9 2nd instar to 3 rd 91.4 97.7 104.4 93.4 ± 8.0 3rd instar to wandering 135.1 148.0 161.4 150.0 ± 10.0 Wandering to puparium 162.5 175.5 189.1 176.6 ± 11.2 Puparium to adult 333.5 359.2 385.9 347.2 ± 23.6

PAGE 65

65 Table 4 3. Cumulative times from oviposition to each life stage transition for Chrysomya megacephala at the 10 th , 50 th (median) and 90 th percentiles (hours) and mean transition times (hours ± SD) at 25.8 °C constant tempe rature. Percentile values were back transformed from square root. Quantile values were predicted from modeling res ults, and means were calculated from laboratory data. Stage 10th 50th 90th Mean ± SD Egg to 1st instar 12.3 14.3 16.5 13.9 ± 0.6 1st instar to 2 nd 31.1 35.0 39.3 33.9 ± 1.6 2nd instar to 3 rd 51.1 55.9 61.0 54.4 ± 4.0 3rd instar to wandering 77.4 87.1 97.5 85.8 ± 4.5 Wandering to puparium 99.3 109.5 120.3 108.0 ± 7.4 Puparium to adult 205.6 225.9 247.1 221.1 ± 4.7 Table 4 4. Cumulative times from oviposition to each life stage transition for Chrysomya megacephala at the 10 th , 50 th (median) and 90 th percentiles (hours) and mean transition times (hours ± SD) at 30.8 °C constant temperature. Percentile values were back transformed from square root. Quantile values were predicted from modeling results, and means were calculated from laboratory data. S tage 10th 50th 90th Mean ± SD Egg to 1st instar 8.5 10.2 12.1 11.1 ± 2.6 1st instar to 2 nd 20.2 23.4 26.9 24.1 ± 3.2 2nd instar to 3 rd 33.4 37.3 41.5 37.5 ± 5.6 3rd instar to wandering 60.8 69.5 78.9 64.0 ± 9.6 Wandering to puparium 81.1 90.4 100.2 86.8 ± 9.1 Puparium to adult 156.6 174.3 193.1 176.8 ± 15.6 Table 4 5. Cumulative times from oviposition to each life stage transition for Chrysomya megacephala at the 10 th , 50 th (median) and 90 th percentiles (hours) and mean transition times (hours ± SD) at 35.6 °C constant temperature. Percentile values were back transformed from square root. Quantile values were predicted from modeling results, and means were calculated from laboratory data. Sta ge 10th 50th 90th Mean ± SD Egg to 1st instar 8.5 10.1 12.0 9.2 ± 2.0 1st instar to 2 nd 19.3 22.5 25.9 22.6 ± 2.9 2nd instar to 3 rd 29.1 32.8 36.7 32.8 ± 1.9 3rd instar to wandering 50.8 58.8 67.3 55.7 ± 5.2 Wandering to puparium 78.3 87.5 97.2 80.9 ± 4.9 Puparium to adult 151.8 169.2 187.7 164.7 ± 8.1

PAGE 66

66 Table 4 6. Predicted error in development time (h) from oviposition to adult emergence, per 1.0 °C error in environmental chamber temperature. Set Temperature (°C) Error (h per 1 °C) 15.0 52 20.0 31 25.0 9 30.0 1

PAGE 67

67 Table 4 7. Mean and peak l arval mass temperatures from pig carcasses in field trials conducted in Earleton, FL, from 2002 2004. Initial ambient temp. (°C) Number of pigs (N) Mean (± SD) day that mass temps. elevated above ambient Mean maximum mass temp. (°C) Peak maggot mass temp. (°C) Number of days peak maximum temps. attained (d) Temp. increase over ambient (%) 14.1 17.5 12 4.2 ± 2.4 38.3 43.2 7 142 17.6 22.5 3 3.6 ± 2.9 39 42.6 4 95 22.6 27.5 12 3.4 ± 2.5 39 50.4 4 56 27.6 32.5 7 1.8 ± .8 41.9 45.1 4 40 32.6 34.5 4 2.3 ± .6 44.6 45.1 4 33

PAGE 68

68 CHAPTER 5 COMPARISON OF THREE PHENOLOGICAL MODELS FOR BEST FIT OF DEVELOPMENT TIMES FOR THE ORIENTAL LATRINE FLY, CHRYSOMYA MEGACEPHALA (DIPTERA: CALLIPHORIDAE) Introduction One of the most important pieces of information that a forensic entomologist can provide is an estimate of the postmortem interval (PMI), or the time that has elapsed since a person died. This is often done by determi ning the age of the calliphorid specimens collected from a corpse. Calliphorids are usually the first insects to arrive at a corpse in a predictable manner. In a study by Kashyap and Pillay (1989), it was determined that the entomological evidence provided the best PMI estimates, better than autopsy reports or eyewitness accounts. Age of the specimens can be estimated using a development (phenological) model with insect development data sets and with weather data obtained from airports located near the deat h scene. I nsects are poikilotherms and their rate of development is a function of temperature . T he relationship between temperatures and insect development rate s i s nonlinear at low and high temperatures, and linear in the intermediate temperatur e range (Arnold 1959, Wagner et al. 1984) . According to Higley and Haskell (2001) , there generally are two approaches for measuring insect development using mathematical models; linear and curvilinear. These models try to predict mathematically the rate of development of a particular species based mainly on temperature. Some of the mod els may be difficult to use because parameter estimation requires learning how to use complicated statistical software programs. Parameter estimation may be time consuming. In a publication regarding linear and nonlinear modeling, Karimi Malati et al. (2014) indicated that one of

PAGE 69

69 the nonlinear model parameter estimations took three hours of iterations on a personal computer. The objective of this study was to determine the best fit of Chrysomya megacephala development data with three wel l known models: the degree day model, the Schoolfield et al. (1995) modification of the Sharpe and DeMichele model (1977) , and a Lactin et al. (1995) modification of the Logan et al. (1976) model. Linear Model (Degree Day M odel) Originally, degree day models were useful for underst anding insect and plant phenology (Higley et al. 1986) in plant science, pest management and ecology. For example, the degree day model might help a farmer calculate the best timing for pesticide applications. According to Wagner et al. (1984), the degree day concept is based on the work of several authors dating back to the mid 1800s. The degree day model assumes that a fixed quantity of metabolic activity , under the control of time and temperature , is required to complete an insect dev elopment. According to the model , the amount of heat required to complete development does not vary, nor will it change due to fluctuating or cons tant temperatures. The model assumptions are: 1. A constant increase in development rate with increasing temperat ure. 2. No development occurs below the minimum developmental threshold. 3. The development rate changes at or above the maximum threshold , though there are different methods of treating this change . The degree day model is: y = m x + b

PAGE 70

70 where y = 1/ T (develop ment time in days), x is the minimum threshold temperature, m and b are fitted parameters determined by linear regression. Determination of developmental thresholds and degree day accumulations over time are prerequisites for use of the degree day mode l . Estimation of the developmental threshold temperature (the x intercept) is determined by extrapolating back the straight l ine plot where the development rate is zero (Higley and Haskell 2001). The Modified Sharpe and DeMichele Model (Schoolfield et al. 1981) A physiological model of poikilotherm development was developed by Sharpe and DeMiche le (1977). The temperature dependent rate model is based upon highly complex equations by Eyring (1935) , Johnson and Lewin (1946) and Hultin (1955) . The Sharpe and DeMichele (1975) model has the following consequences: 1. It demonstrates the validity of the linear approximation (degree day concept) in the mid temperature region for some organisms. 2. It effectively establishes a low temperature threshold for development 3. It reduces the rate of development at higher temperatures, thereby establishing both an optimum and upper threshold for development. Sharpe and DeMichele (1975) identified therm odynamic constants that , including two species of insects: Pseud atomoscelis seriatus Reuter (cotton fleahopper) and Drosophila melanogaster Meigen (fruit fly). The model fit the development rates well. But the Sharpe and DeMichele model was not well suited for non linear regression and was subsequently modified by Sch oolfield et al. (1981) . The modification

PAGE 71

71 alleviated the non linear regression problem and allowed more convenient initial parameter estimation. For cases with limited data availabi lity at higher temperatures, a reduced four parameter model was proposed (equation 6 in Schoolfield et al. 1981): where r (T) is the mean development rate at temperature T (time 1 ), T is temperature in Kelvin (298 k = 25 °C), , , , and are fitted parameters. The Modified L ogan et al. Model (Lactin et al. 1995) L ogan et al. (1976) developed a nonlinear model to describe growth rates of arthropods to address th e shortcomings of the degree day model outside of the linear range of temperatures. L actin et al. (1995) modified the Logan et al. (1976 ) model by (modification 2), that allows for estimation of a developmental threshold temperature. Lactin model, but recommended inclusion of this parameter when temperatures are near the low developmental threshold. The Lactin et al. (1995) 2 modification is: max (T max max , and parameters. Voss et al. (2014) recommended use of the Lactin modification 2 for calculation of degree d ays for development of the forensically important Calliphoridae, Calliphora varifrons Malloch. The objective of this study was to determine the best fit of C . megacephala development data with three well known models: the degree day model, a Schoolfield et

PAGE 72

72 al. (1995) modification of the Sharpe and DeMichele model (1977) , and a Lactin et al. (1995) modification of the Logan et al. (1976) model. Materials and Methods Models and variations of models as described above were fit to the development rate data set from laboratory reared C . megacephala ( Chapter 4) . The models compared were: Model 1. Degree day model with complete data set (16.0°C 35.6 °C) Model 2. Degree day model with incomplete data set (16.0°C 30.8 °C) Model 3. Degree day model with incomplete data set and 10 °C minim um threshold Model 4. Degree day model with incomplete data set and 11°C minimum threshold Model 5. Lactin modification 2 of Logan model with x Model 6. Lactin modification 2 of Logan model without x Model 7. Schoolfield 4 parameter modification of Sharpe and DeMichele model M odel s 1 4 are linear models, 5 7 are nonlinear. Model 1 was fit ted with the full data set. Linear models 2 4 were fitted to , and compared with a reduced data set, using only development data from 16. 1 30.8 °C . The 10 °C development threshold temperature used in model 3 was chosen because it is commonly used in case reports; the 11 °C minimum threshold temperature used in model 4 was calculated by extrapolation. Initial parameters were chosen from pub lished parameter values of similar species (Schoolfield et al. 1981, Lactin et al. 1995) . The nls procedure in R was used to fit the nonlinear models (Baty and Delignette Muller 2013) . Output from the procedure include d the final fitted parameters and the Residual Sum of Squares (RSS) error of the fitted model; thus , allowing for comparison amon g models for goodness of fit. A lower RSS indicates a better fit.

PAGE 73

73 Results Models with estimated parameter values and RSS are in Table 5 1 . C hrysomya megacephala development times are fitted in linear models 1 and 4, and in nonlinear models 5, 6 and 7 (Figure 5 1). T he linear model performed better for temperatures in the intermediate temperature pa rt of the development curve (16 30.8 °C). The nonlinear models (models 5 7) fit the complete (16.0 35.6 °C) C . m egacephala data set well, but Model 7 had the best fit. The full model (equation 4 in Schoolfield et al. 1981) did not converge well with the C. megacephala data set, so the reduced model was used (equation 6), as recommended by Schoolfield et al. (1981) w hen development rates at high temperatures are unknown. Discussion One of the prerequisites for using the degree day model is that the minimum threshold temperature be known. Few calculated estimates of developmental minima or maxima exist for forensically important calliphorids. M inimum threshold temperatures usually are determined imprecisely because determination of t he x intercept with extrapolation is only an approximation . Higley and Haskell (2001) estimated that forensically important flies have a de velopmental threshold temperature that is between 6 and 10 °C. Use of 10 °C as the minimum threshold temperature for degree day calculations is commonly used in case reports (VanLaerhoven 2008) . VanLaerhoven (2008) conducted the only blind validation study to date, using three pig s and Phormia regina data sets from Kamal (1958) , Nishida (1984) , Greenberg (1991) , Anderson (2000) and Byrd and Allen (2001) . To determine a PMI, she used three different minimum threshold temperatures: 0, 6, and 10 °C, and found that

PAGE 74

74 increasing the minimum threshold temperatures could add error to the PMI estimations, depending on the data set used. The degree da y model may be appropriate to predict larval emergence of insects such as the olive psyllid Euophyllura phillyreae Foerster (Kumral et al. 2008) , the armored scale, Melanaspis deklei Deitz and Davidson (Chong and Camacho 2014) , and the leaf mining fly, Hydrellis lagarosiphon Deeming (Mangan and Baars 2013) but larvae of these insects has behavior and physiology similar to the calliphorids (i.e. , large masses that generate heat ). The degree day model is widely used because it is easy to use, easy to apply , and yields approximately correct values, but it has limitations that need to be taken into consideration . Pruess (1983) noted that there have not been many attempts to validate degree day models , and this is especially true in the field of forensic entomol ogy . For forensically important calliphorids, there are few data sets that include temperature extremes, or even a range of temperatures. In a field study by Slone and Gruner (2007) , it was determined that large volumes of larvae had mean elevated temperatures higher than 38 °C, even when the ambient temperatures were only 15 20 °C. At ambient temperatures from 15 40 °C, large volumes of larvae were capable of generating peak temperatures over 43 °C, and could attain high te mperatures for days (Table 4 7). Similarly, in a laboratory study by Gruner it was found that volumes > 50 cm 3 could generate heat higher than 31 °C at ambient temperatures of only 21.2 °C, and that volumes > 150 cm 3 could generate heat over 40 °C at times ( Fig ure 3 3 ). Degree day calculations at these ambient temperatures would

PAGE 75

75 indicate slower development than actually occurred, and would result in an underestimation of accumulated degree days. One of the disadvantages of the nonlinear models is that da ta sets with a wide range of temperatures are needed for the models to work. Calliphorid data sets with additional data are needed to test the fit of these models before an optimal model can be declared. It has been proposed that the degree day model is of calliphorid larval or puparial ages, as long as their development temperatures are in the linear part of the development curve (Higley and Haskell 2001). But use of the degree day model for development temperatures outside t he linear range (here between 15 and 3 0.8 °C ) would add error to any PMI estimation . We documented that maggot mass temperatures frequently exceed 30.8 °C ( Fig ure 4 7 ). As these higher temperatures are outside of the linear range, this calls into question use of linear models for larval development in maggot masses. These results provide further insight for application of curvilinear models to predict age of calliphorid specimens collected from a corpse.

PAGE 76

7 6 Figure 5 1 . Chrysomya megacephala development rates fitted in models 1,4,5,6, and 7.

PAGE 77

77 Table 5 1. Seven development models and their estimated parameters values. A lower RSS indicates a better fit. Model # Estimated parameter values RSS 1 m = 0.0002557 3.50E 05 b = 0.002393 2 m = 0.0002899 1.71E 05 b = 0.003190 3 m = 0.0002739 1.82E 05 4 m = 0.0002899 1.71E 05 5 1.81E 05 Tmax = 47.499535 0.005138 6 2.24E 05 T max = 41.009674 7 = 0.008137 1.77E 05 A = 3725 L = 35580 T ½ L = 297.6

PAGE 78

78 CHAPTER 6 CONCLUSIONS To date, this is the most extensive study of Chrysomya megacephala development. During preliminary development tests , it was discovered that the liver feeding substrate dried up quickly, resulting in desiccation of the larvae ; thus, a liver agar feeding substrate was developed that did not desiccate as quickly as liver. This substrate can be used for feeding other s pecies of calliphorids. It is well known that maggot masses gener ate heat. Several authors reported that elevated maggot mass temperatures were determined by larval stage (instar) or number of instars in a mass. As the number and instar of larvae in a mass are correlated with the volume of the mass larval instars, they cannot both be included as predictors in a traditional general linear mixed model (GLMM). Twelve candidate multiple regression models were fitted to the data; the models were then scored with Akaike information criterion (AIC) and ranked with the Bayesian information criterion (BIC). The results of this study indicate that although instar, age, treatment temperature, and number of maggots in a mass were significant, larval volume was the best single predictor of maggot mass temperatures , but the most parsi monious model included larval volume and treatment temperature. This is consistent with the results of the field study by Slone and Gruner ( 2007) , in which it was determined that volume was the principal predictor of maggot mass temperatures in field depos ited pigs . The m ajor objective of this research was to obtain development and population transition times for C. megacephala at constant temperatures. As larvae in very small masses are unable to thermoregulate , the effect of the maggot mass temperatures on

PAGE 79

79 development time was eliminated by using only 10 larvae per cup. Eggs did not survive at 5, 10 and 40 °C; thus, development times were not obtained at these temperatures. T he development and transition times of C. megacephala were determined at five constant temperatures: 16, 21.2, 25.8, 30.8 and 35.6 °C. Survivorship at each temperature was 14% (16 °C), 83% (21.2 °C), 85% (25.8 °C), 67% (30.8 °C) and 55% (35.6 °C). The larvae developed fastest at 35.6 °C. Bas ed on survivorship, the optimal growth temperature for C. megacephala may be about 25.0 °C ambient temperature. More research is required on development at a wide range of temperatures, especially at high temperatures over 35 ºC. One of the most important pieces of information that a forensic entomologist can estimate is the postmortem interval (PMI), or the time that has elapsed since a person died. This is often done by determining the age of the calliphorid specimens collected from a corpse. Age of the specimens can be estimated using a development (phenological) model with insect development data sets and with weather data obtained from airports located near the death scene. The degree day linear model was compared to two commonly used nonlinear pheno logical models. The nonlinear models fit the complete C . megacephala data set (16.0 35.6 °C) better than did the linear models The degree day model is determination of calliphorid larval or puparial ages, as long as their development tempera tures are in the linear part of the development curve bu t use of the degree day model for development temperatures outside the linear range would add error to any PMI estimation . And m aggot mass temperatures frequently exceed 38 .0 °C ( Figure 4 7)

PAGE 80

80 which is outside of the linear range of development. This calls into question the use of linear models for development of larvae in maggot masses. The results of this research provide further insight for application of nonlinear models to predict age of calliphorid specimens collected from a corpse. More research is needed to test the nonlinear models with high temperatures that calliphorid larvae generate.

PAGE 81

81 APPENDIX A R CODES FOR CHAPTER 3 require(lme4) DATA=read.csv( file="volume.csv", header = TRUE) DATA$Ftemp = factor(DATA$Temp.Trt) m1 < lmer(log.mass.temp ~ log.num + (1|CID) , data=DATA) m2 < lmer(log.mass.temp ~ instar + (1|CID) , data=DATA) m3 < lmer(log.mass.temp ~ Ftemp + (1|CID) , data=DATA) m4 < lmer( log.mass.temp ~ poly(log.vol,2) + (1|CID) , data=DATA) m5 < lmer(log.mass.temp ~ log.num * Ftemp + (1|CID) , data=DATA) m6 < lmer(log.mass.temp ~ instar * Ftemp + (1|CID) , data=DATA) m7 < lmer(log.mass.temp ~ log.num * instar + (1|CID) , data=DATA) m8 < lmer(log.mass.temp ~ Ftemp * poly(elapsed,2) + (1|CID) , data=DATA) m9 < lmer(log.mass.temp ~ Ftemp * poly(log.vol,2) + (1|CID) , data=DATA) m10< lmer(log.mass.temp ~ log.num * instar * Ftemp + (1|CID) , data=DATA) m11< lmer(log.mass.temp ~ pol y(log.vol,2) * Ftemp * poly(elapsed,2) + (1|CID) , data=DATA) m12< lmer(log.mass.temp ~ log.num * instar * Ftemp * poly(elapsed,2) + (1|CID) , data=DATA) anova(m1,m2,m3,m4,m5,m6,m7,m8,m9,m10,m11,m12)

PAGE 82

82 A PPENDIX B TIMES TO TRANSITION (H) AT THE 10TH, 50T H, AND 90TH PERCENTI LE ± SE

PAGE 83

83 Temp. (°C) Stage transition 10 SE 10th % 10 SE+ 50 SE Median 50 SE+ 90 SE 90th % 90 SE+ 16.0 Egg to instar 1 43.0 43.7 44.9 46.6 47.5 48.3 50.2 51.4 52.2 Instar 1 to 2 119.5 121.0 123.1 127.1 128.7 130.4 134.6 136.8 138.4 Instar 2 to 3 199.2 201.0 203.3 208.6 210.4 212.3 217.6 220.1 221.8 Instar 3 to wandering 325.2 329.2 335.1 344.5 349.1 353.7 363.5 369.6 373.9 Wandering to puparium 376.3 380.7 387.8 395.4 400.6 405.8 413.6 421.0 425.6 Puparium to adult 725.5 731.1 738.5 762.4 768.9 775.3 799.2 807.6 813.9 21.2 Egg to instar 1 20.9 21.4 22.0 23.5 24.0 24.5 26.1 26.8 27.4 Instar 1 to 2 55.2 56.2 57.5 60.4 61.5 62.6 65.6 67.1 68.2 Instar 2 to 3 90.2 91.4 93.0 96.3 97.7 99.0 102.6 104.4 105.6 Instar 3 to wandering 132.4 135.1 139.3 144.8 148.0 151.2 156.9 161.4 164.4 Wandering to puparium 159.5 162.5 167.1 172.1 175.5 179.0 184.2 189.1 192.1 Puparium to adult 330.9 333.5 336.7 356.4 359.2 362.1 382.1 385.9 388.9 25.8 Egg to instar 1 12.0 12.3 12.6 14.0 14.3 14.6 16.0 16.5 16.8 Instar 1 to 2 30.5 31.1 31.7 34.5 35.0 35.6 38.5 39.3 39.9 Instar 2 to 3 50.4 51.1 51.9 55.2 55.9 56.6 60.1 61.0 61.7 Instar 3 to wandering 75.9 77.4 79.1 85.7 87.1 88.6 95.5 97.5 99.1 Wandering to puparium 97.4 99.3 101.9 107.4 109.5 111.6 117.4 120.3 122.3 Puparium to adult 204.0 205.6 207.4 224.3 225.9 227.5 244.8 247.1 249.1 30.8 Egg to instar 1 8.3 8.5 8.9 9.9 10.2 10.5 11.6 12.1 12.4 Instar 1 to 2 19.6 20.2 20.8 22.8 23.4 24.0 26.1 26.9 27.5 Instar 2 to 3 32.8 33.4 34.3 36.6 37.3 37.9 40.6 41.5 42.2 Instar 3 to wandering 59.3 60.8 62.9 67.8 69.5 71.2 76.7 78.9 80.4 Wandering to puparium 79.4 81.1 83.6 88.4 90.4 92.4 97.5 100.2 102.1 Puparium to adult 155.0 156.6 158.5 172.4 174.3 176.0 190.6 193.1 195.1 35.6 Egg to instar 1 8.2 8.5 8.9 9.7 10.1 10.4 11.5 12.0 12.4 Instar 1 to 2 18.8 19.3 20.1 21.9 22.5 23.2 25.1 25.9 26.6 Instar 2 to 3 28.5 29.1 30.0 32.1 32.8 33.6 35.6 36.7 37.4 Instar 3 to wandering 49.3 50.8 52.9 57.1 58.8 60.5 65.0 67.3 69.0 Wandering to puparium 76.7 78.3 80.8 85.5 87.5 89.5 94.5 97.2 99.1 Puparium to adult 149.9 151.8 154.0 167.1 169.2 171.3 184.8 187.7 190.0

PAGE 84

84 APPENDIX C R CODE FOR DEVELOPMENT AL POPULATION TRANSI TION TIMES DATA=read.csv(file="DevelopmentRates.csv", header = TRUE) choosestage < 1 # which transition to model DATA$stage1< (DATA$stage>=choosestage)*1 mylogit1 < glm(stage1 ~ sqrt(Elapsed) + factor(Temp.Trt) + factor(Unit) + factor(CID) + factor(Trial), data = DATA, family = "binomial") newdata1 < with(mydata, data.frame(Elapsed = seq(from = .1, to = 2000, by=1), Unit. = 39,Cup. = 1,Trial. =1)) newdata1 < with(DATA, data.frame(Elapsed = rep(seq(from = 1, to = 1001, length.out = 10001), 5), Unit. = 39,Cup. = 1,Trial. =1, Temp.Trt = factor(rep(1:5, each = 10001)))) newdata1 < cbind(newdata1, predict(mylogit1, newdata = newdata1, type = "response", se = TRUE)) require(lme4) DATA=read.csv(file="DevelopmentRates.csv", header = TRUE) DATA < subset(DATA,Tran sition == 6)

PAGE 85

85 APPENDIX D R CODE S FOR MODEL PARAMETER ESTIMATION DEGREE DAY MODEL DATA.subset < subset(DATA,Temp <35) dd1all < nls( rate~m*Temp2+b,DATA, start = list(m = .0002,b= .003)) dd1 < nls( rate~m*Temp2+b,DATA.subset, start = list(m = .0002,b= .003)) dd2 < nls( rate~m*(Temp2 10),DATA.subset, start = list(m = .0002)) dd3 < nls( rate~m*(Temp2 11),DATA.subset, start = list(m = .0002)) LACTIN MODEL L1 < nls( rate~exp(rho*Temp2) exp(rho*Tmax (Tmax Temp2)/delta)+lambda,D ATA, start = list(rho=.1,Tmax=45,delta=7,lambda= .02),trace=TRUE ,control=list(maxiter=20000,tol=1e 4,minFactor=1e 100)) L2 < nls( rate~exp(rho*Temp2) exp(rho*Tmax (Tmax Temp2)/delta),DATA, start = list(rho=.1,Tmax=45,delta=7),trace=TRUE ,control=list( maxiter=20000,tol=1e 4,minFactor=1e 10,printEval=TRUE)) SCHOOLFIELD MODEL S1 < nls( rate~(rho*((Temp2+273)/298)*exp((DHA/1.987)*(1/298 1/(Temp2+273)) ))/(1+exp(DHL/1.987*(1/T12L 1/(Temp2+273))) ),DATA,start = list(rho=.273,DHA=9963,DHL= 51510,T12L=291. 2),trace=TRUE ,control=list(maxiter=20000,tol=1e 4,minFactor=1e 100,printEval=TRUE))

PAGE 86

86 LIST OF REFERENCES Ames, C., and B. Turner. 2003. Low temperature episodes in development of blowflies: implications for postmortem interval estimation. Medical and Ve terinary Entomology 17: 178 186. Anderson, G. S. 2000. Minimum and maximum developmental rates of some forensically important Calliphoridae (Diptera). Journal of Forensic Sciences 45: 824 832. Anderson, G. S. 2001a. Forensic entomology in British Columbia: A brief history. Journal of the Entomological Society of British Columbia 98: 127 135. Anderson, G. S. 2001b . Insect succession on carrion and its relationship to determining time of death, pp. 143 175. In J. H . Byrd and J. L. Castner [eds.], Forensic Entomology: The Utility of Arthropods in Legal Investigations. CRC Press, Boca Raton. Anderson, G. S., and S. L. VanLaerhoven. 1996. Initial studies on insect succession on carrion in southwestern British Columbia. Journal of Forensic Sciences 41: 617 625. Arnold, C. Y. 1959. The determination and significance of the base temperature in a linear heat unit system. Proceedings of the American Society for Horticultural Scienc e 74: 430 445. Bates, D., M. Maechler, B. Bolker, and S. Walker. 2014. lmer4: Linear mixed effects models using Eigen and S4. P package version 1.1 6. http://CRAN.R project.org/package=lme4 . Baty, F., and M. L. Delignette Muller 2013. nlstools: Tools for nonlinear regression diagnostics computer program, version R package version 0.0 15. Baumgartner, D. L. 1988. Spread of introduced Chrysomya blowflies (Diptera: Calliphoridae) in the Neotropics with record new to Venezuela. Biotropica 20: 167 168. Benecke, M. 1998. Six forensic entomology cases: description and commentary. Bohart, G. E., and J. L. Gressitt. 1951. Filth inhabiting flies of Guam. Bernice P. Bishop Museum Bulletin Honolulu. Bull. 204. 169 pp.

PAGE 87

87 Bradshaw, J. E., D. J. Gemmell, S. Gowers, and R. N. Wilson. 2002. Turnip ( Brassica rapa L. ssp. rapifera Metzg.) population improvement and cultivar production. Plant Breeding 121: 301 306. Burnham, K. P., and D. R. Anderson. 2001. Kullback Leibler information as a basis for strong inference in ecological studies. Wildlife Research 28: 111 119. Burnham, K. P., and D. R. Anderson. 2004. Multimodel inference: Understanding AIC and BIC in model selection. Sociological Methods and Research 33: 261 304. Byrd, J. H., and J. C. Allen. 2001 . The development of the black blow fly, Phormia regina . Forensic Science International 120: 79 88. Byrd, J. H., and J. F. Butler. 1997. Effects of temperature on Chrysomya rufifacies (Diptera: Calliphoridae) development. Journal of Medical Entomology 34: 353 357. Campobasso, C. P., G. Di Vella, and F. Introna. 2001. Factors affecting decomposition and Diptera colonization. Forensic Science International 120: 18 27. Catts, E. P., and N. H. Haskell. 1990. Entomology & Death A Procedural Guide. Joyce's Print Shop, Inc., Clemson, SC. Charabidze, D., B. Bourel, and D. Gosset. 2011. Larval mass effect: Characterisation of heat emission by necrophageous blowflies (Diptera: Calliphoridae) larval aggregates. Forensic Science International 211: 61 66. Chong, J. H., and E. R. Camacho. 2014. Distribution, host plants, and life cycle of Melanaspis deklei ( Hemiptera: Coccoidea: Diaspididiae) in South Carolina, USA. Acta Zoologica Bulgarica Supplementum 6 (2014): 13 19. Cianci, T. J., and J. K. Sheldon. 1990. Endothermic generation by blow fly larvae Phormia regina developing in pig carcasses. Bulletin of the Society of Vector Ecology 15: 33 40. Cragg, J. B. 1956. The olfactory behavior of Lucilia species (Diptera) under natural conditions. Annals of Applied Biology 44: 467 477. Deonier, C. C. 1940 . Carcass temperatures and their relation to winter b lowfly populations and activity in the southwest. Journal of Economic Entomology 33: 166 170. Donovan, S. E., M. J. R. Hall, B. D. Turner, and C. B. Moncrieff. 2006 . Larval growth rates of the blowfly, Calliphora vicina , over a range of temperatures. Medical and Veterinary Entomology 20: 106 114.

PAGE 88

88 Early, M., and M. L. Goff. 1986. Arthropod succession patterns in exposed carrion on the island of O'ahu, Hawaiian Islands, USA. Journal of Medical Entomology 23: 520 531. Easton, A. M., and K. G. V. Smith. 1970. The entomology of the cadaver. Medicine, Science and the Law 10: 208 219. Esser, J. R. 1991. Biology of Chrysomya megacephala (Diptera: Calliphoridae) and reduction of losses caused to the salted dried fish industry in south east Asia. Bullet in of Entomological Research 81: 33 42. Eyring, H. 1935. The activated complex in chemical reactions. The Journal of Chemical Physics 3: 107 115. Gabre, R. M., F. K. Adham, and H. Chi. 2005. Life table of Chrysomya megacephala (Fabricius) (Diptera : Calliphoridae). Acta Oecologica 27: 179 183. Gallagher, M. B., S. Sandhu, and R. Kimsey. 2010. Variation in developmental time for geographically distinct populations of the common green bottle fly, Lucilia sericata (Meigen). Journal of Forensic Sciences 55: 438 442. Gilead, L., K. Y. Mumcuoglu, and A. Ingber. 2012. The use of maggot debridement therapy in the treatment of chronic wounds in hospitalised and ambulatory patients. Journal of Wound Care 21: 78, 80, 82 85. Goff, M. L. 1991. Compar ison of insect species associated with decomposing remains recovered inside dwellings and outdoors on the island of Oahu, Hawaii. Journal of Forensic Sciences 36: 748 753. Goodbrod, J. R., and M. L. Goff. 1990. Effects of larval population density on rates of development and interactions between two species of Chrysomya (Diptera: Calliphoridae) in laboratory culture. Journal of Medical Entomology 27: 338 343. Grassberger, M., and C. Reiter. 2001. Effect on temperature on Lucilia sericata (Diptera: Calliphoridae) development with special reference to the isomegalen and isomorphen diagram. Forensic Science International 120: 32 36. Greenberg, B. 1991. Flies as forensic indicators. Journal of Me dical Entomology 28: 565 577. Greenberg, B. 1988. Chrysomya megacephala (F.) (Diptera: Calliphoridae) collected in North America and notes on Chrysomya species present in the New World. Journal of Medical Entomology 25: 199 200.

PAGE 89

89 Greenberg, B., and J. C. Kunich. 2002. Entomology and the Law: Flies as Forensic Indicators. Cambridge University Press, Cambridge. Gruner, S. V., and D. H. Slone. 2014. A fresh liver agar substrate for rearing small numbers of forensically important blow flies (Diptera: Calliphoridae). Journal of Medical Entomology 51: 713 715. Gruner, S. V. 2008. Maggot Therapy, pp. 2257 2258. In J. L. Capinera [ed.], Encyclopedia of Entomology, Second ed. Springer Netherlands, Boston. Gruner, S. V., D. H. Slone, and J. L. Capinera. 2007. Forensically important Calliphoridae (Diptera) associated with pig carrion in rural north central Florida. Journal of Medical Entomology 44: 509 515. Hall, D. G. 1948. The Blowflies of North America. Thomas Say Foundation, Lafayette, IN. 477 pp. Heaton, V., C. Moffatt, and T. Simmons. 2014. Quantifying the temperature of maggot masses and its relationship to decomposition. Journal of Forensic Sciences 59: 676 682. Higley, L. G., and N. H. Haskell. 2001. Insect development and forensic entomology. In J. H. Byrd and J. L. Castner [eds.], Forensic Entomology: The utility of art hropods in legal investigations. CRC Press, Boca Raton. Higley, L. G., L. P. Pedigo, and K. R. Ostlie. 1986. DEGDAY: A program for calculating degree days, and assumptions behind the degree day approach. Environmental Entomology 15: 999 1016. Hultin, E. 1955. The influence of temperature on the rate of enzymic processes. Acta Chemica Scandinavica 9: 1700 1710. Huntington, T. E., L. G. Higley, and F. P. Baxendale. 2007. Maggot development during morgue storage and its effect on estimating the post mortem interval. Journal of Forensic Sciences 52: 453 458. Introna, F., Jr., C. P. Campobasso, and A. Di Fazio. 1998. Three case studies in forensic entomology from southern Italy. Journal of Forensic Sciences 43: 210 214. Introna , F., Jr., T. W. Suman, and J. E. Smialek. 1991. Sarcosaprophagous fly activity in Maryland. Journal of Forensic Sciences 36: 238 243.

PAGE 90

90 Ireland, S., and B. Turner. 2005. The effects of larval crowding and food type on the size and development of the blowfly, Calliphora vomitoria . Forensic Science International 159: 175 181. James, M. T. 1947. The Flies That Cause Myiasis in Man, pp. 175, Miscellaneous Publication No. 631. United States Department of Agriculture, Washington, D.C. Johnson, F. H., and I. Lewin. 1946. The disinfection of E. coli in relation to temperature, hydrostatic pressure and quinine. Journal of Cellular and Comparative Physiology 28: 23 45. Joy, J. E., N. L. Liette, and H. L. Harrah. 2006. Carrion fly (Diptera: Calliphoridae) larval colonization of sunlit and shaded pig carcasses in West Virginia, USA. Forensic Science International 164: 183 192. Kamal, A. S. 1958. Comparative study of thirteen species of sarcosaprophagous Calliphoridae and Sarcophagidae (Diptera) I. Bionomics. Annals of the Entomological Society of America 51: 261 271. Karimi Malati, A., Y. Fathipour, and A. A. Talebi. 2014 . Development response of Spodoptera exigua to eight constant temperatures: Linear and nonlinear mo deling. Journal of Asia Pacific Entomology 17: 349 354. Kashyap, V. K., and V. V. Pillay. 1989. Efficacy of entomological method in estimation of postmortem interval: A comparative analysis. Forensic Science International 40: 245 250. Keh, B. 1985 . Scope and applications of forensic entomology. Annual Review of Entomology 30: 137 154. Kerridge, A., H. Lappin Scott, and J. R. Stevens. 2005 . Antibacterial properties of larval secretions of the blowfly, Lucilia sericata . Medical and Veterinary Entomology 19: 333 337. Khole, V. 1979. Studies on metabolism in relation to post embryonic development of s ome calliphorid flies (Diptera: Insecta). Entomon 4: 61 63. Kumral, N. A., B. Kovanci, and B. Akbudak. 2008. Using degree day accumulations and host phenology for predicting larval emergence patterns of the olive psyllid, Euphyllura phillyreae . Journal of Pest Science 81: 63 69. Lactin, D., J., N. J. Holliday, D. L. Johnson, and R. Craigen. 1995 . Improved rate model of temperature dependent development by arthropods. Environmental Entomology 24: 68 75.

PAGE 91

91 Lane, R. P. 1975. An investigation into blowfly (Diptera: Calliphoridae) succession on corpses. Journal of Natural History 9: 581 598. Lehneback, C. A., and A. W. Robertson. 2004 . Pollination ecology of four epiphytic orchids of New Zealand. Annals of Botany 93: 773 781. Logan, J. A., D. J. Wollkind, S. C. Hoyt, and L. K. Tanigoshi. 1976 . An analytic model for description of temperature dependent rate phenomenon in arthropods. Environmental Entomology 5: 1133 1140. Mangan, R., and J. R. Baars. 2013. Use of life table statistics and degree day values to predict the colonisation success of Hydrellia lagarosi phon Deeming (Diptera: Ephydridae), a leaf mining fly of Lagarosiphon major (Ridley) Moss (Hydrocharitaceae), in Ireland and the rest of Europe. Biological Control 64: 143 151. Marchenko, M. I. 2001. Medicolegal relevance of cadaver entomofauna for the determination of the time of death. Forensic Science International 120: 89 109. Martinez Sanchez, A., M. S. Marcos Garcia, and S. Rojo. 2001. First collection of Chrysomya megacephala (Fabr.) in Europe (Diptera: Calliphoridae) Pan Pacific Entomologist 77: 240 243. Nabity, P. D., L. G. Higley, and T. M. Heng Moss. 2006. Effects of temperature on development of Phormia regina (Diptera: Calliphoridae) and use of developmental data in determining time intervals in forensic entomology. Journal of Medical Entomology 43: 1276 1286. Nelson, L. A., M. Dowton, and J. F. Wallman. 2009. Thermal attributes of Chrysomya species. Entomologia Experimentali s et Applicata 133: 260 275. Nigam, Y., A. Bexfield, S. Thomas, and N. A. Ratcliffe. 2006. Maggot therapy: T he science and implication for CAM Part II maggots combat infection. Evidenced based Complementary and Alternative Medicine 3: 303 308. Nishida, K. 1984. Experimental studies on the estimation of postmortem intervals by means of fly larvae infesting human cadavers. Japanese Journal of Legal Medicine 38: 24 41. Nuorteva, P. 1977. Sarcosaprophagous insects as forensic indicators, pp. 1072 1095. In W. B. S. Company [ed.], Forensic Medicine: A Study in Trauma and Environmental Hazards, Toronto. Nuorteva, P., M. Isokoski, and K. Laiho. 1967. Studies on the possibilities of using blowflies (Dipt.) as medicolegal indicators in Finland. Annals of Entomological Fennica 33: 217 225.

PAGE 92

92 O'Flynn, M. A. 1983. The succession and rate of development of blowflies in carrion i n southern Queensland and the application of these data to forensic entomology. Journal of the Australian Entomological Society 22: 137 148. Payne, J. A. 1965. A summer carrion study of t he baby pig Sus scrofa Linnaeus. Ecology 46: 592 602. Pharr, L. R. 2009. A taphonomic model of concealment: Decomposition and the postmortem interval (PMI) in a 55 gallon barrel, Unpublished M.S. thesis, College of Agriculture. Louisiana State Univers ity, Baton Rouge, LA. 245 pp. Prange, H. D. 1995. Evaporative cooling in insects. Journal of Insect Physiology 42: 493 499. Pruess, K. P. 1983. Day Degree methods for pest management. Environmental Entomology 12: 613 619. Putman, R. J. 1978. The role of carrion frequenting arthropods in the decay process. Ecological Entomology 3: 133 139. Putman, R. J. 1983. Carrion and Dung: the decomposition of animal wastes, pp. 59, The institute of biology's studies in biology no. 156. Edward Arnold (Publishers) Limited, Southhampton. R Core Team (2014). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL < http://w ww.R project.org/ >. Richards, C. S., and M. H. Villet. 200 8 . Factors affecting accuracy and precision of thermal summation models of insect development used to estimate post mortem intervals. International Journal for Legal Medicine 122: 401 408. Richards, C. S., P. W. Price, and M. H. Villet. 2009. Thermal ecophysiology of seven carrion feeding blowflies in Southern Africa. Entomologia Experimentalis et Applicata 131: 11 19. Rivers, D. B., C. Thompson, and R. Brogan. 2011. Physiological trade offs of forming maggot masses by necrophagous flies on vertebrate carrion. Bulletin of Entomological Research 2011: 599 611. Rivers, D. B., T. Ciarlo, M. Spelman, and R. Brogan. 2010. Changes in development and heat shock protein expression in two spe cies of flies ( Sarcophaga bullata [Diptera: Sarcophagidae] and Protophormia terraenovae [Diptera: Calliphoridae]) reared in different sized maggot masses. Journal of Medical Entomology 47: 677 689.

PAGE 93

93 Sarkar, D. 2008. L attice: Multivariate data visuali zation with R. Springer, New York. Schittenhelm, S., T. Giadis, and V. R. Rao. 1997. Efficiency of various insects in germplasm regeneration of carrot, onion and turnip rape accessions. Plant Breeding 116: 369 375. Schoolfield, R. M., P. J. H. Sharpe, and C. E. Magnuson. 1981. Non linear regression of biological temperature dependent rate models based on absolute Reaction rate Theory. Journal of Theoretical Biology: 719 731. Sharov, A. 1998. "Simple degree day model." Quantitive population e cology. 1998. Virginia Polytechnic Institute. November 15, 1998. . Sharpe, P. J. H., and D. W. DeMichele. 1977. Reaction kinetics of poikilotherm development. Journal of Theoretical Biology 64: 649 670. Sherman, R. A., and J. My Tien Tran. 1995. A simple, sterile food source for rearing the larvae of Lucilia sericata (Diptera: Calliphoridae). Medical and Veterinary Entomology 1995: 393 398. Slone, D. H., and S. V. Gruner. 2011. Letter to editor. Journal of Medical Entomology 48: 127 128. Slone, D. H., and S. V. Gruner. 2007. Thermoregulation in larval aggregations of carrion feeding blow flies (Diptera: Calliphoridae). Journal of Medical Entomology 44: 516 523. Stevenson, D. E., G. J. Michels, J. B. Bible, J. A. Jackman, and M. K. Harris. 2008. Physiological time model for predicting adult emergence of western corn rootworm (Coleoptera: Chrysomelidae) in Texas High Plains. Journal of Economic Entomology 101: 154 8 1593. Subramanian, H., and K. R. Mohan. 1980. Biology of the blowflies Chrysomya megacephala , Chrysomya rufifaces and Lucilia cuprina . Kerala Journal of Veterinary Science 11: 252 261. Sukontason, K. L., P. Narongchai, D. Sripakdee, N. Boonchu, T. Chaiwong, R. Ngern Klun, S. Piangjai, and K. Sukontason. 2005. First report of human myiasis caused by Chrysomya megacephala and Chrysomya rufifaces (Diptera: Calliphoridae) in Thailand, and its implication in forensic entomology. Journal of Medical Ento mology 42: 702 704. Sulaiman, S., A. R. Sohadi, and J. Jeffery. 1989. Human helminth parasite burdens on cyclorrhaphan flies (Diptera) trapped at an aboriginal settlement in Malaysia. Bulletin of Entomological Research 79: 625 629.

PAGE 94

94 Sulaiman , S., A. R. Sohadi, H. H. Yunus, and R. R. Iberahim. 1988 . The role of some cyclorrhaphan flies as carriers of human helminths in Malaysia. Medical and Veterinary Entomology 2: 1 6. Tarone, A. M., and D. R. Foran. 2008. Generalized additive models and Lucilia sericata growth: Assessing confidence intervals and error rates in forensic entomology. Journal of Forensic Sciences 53: 942 948. Tomberlin, J. K., W. K. Reeves, and C. Sheppard. 2001. First record of Chrysomya megacephala (Diptera: Calliphori dae) in Georgia, U.S.A. Florida Entomologist 84: 300 301. Toolson, E. C. 1987. Water profligacy as an adaptation to hot deserts water loss rates and evaporative cooling in the Sonoran Desert cicada, Diceroprocta apache (Homoptera: Cicadidae) Physiological Zoology 60: 379 385. Turner, B., and T. Howard. 1992. Metabolic heat generation in dipteran larval aggregations: A consideration for forensic entomology. Medical and Veterinary Entomology 6: 179 181. VanLaerhoven, S. L. 2008. Blind val idation of postmortem interval estimates using developmental rates of blow flies. Forensic Science International 180: 76 80. Voss, S. C., D. F. Cook, W. F. Hung, and I. R. Dadour. 2014. Survival and development of the forensically important blow fly, Calliphora varifrons (Diptera: Calliphoridae) at constant temperatures Forensic Science, Medicine and Pathology 10: 314 321. Wagner, T. L., H. I. Wu, P. J. H. Sharpe, R. M. Schoolfield, and R. M. Coulson. 1984. Modeling insect development rates: A literat ure review and application of a biophysical model. Annals of the Entomological Society of America 77: 208 225. Walker, T. J., J. J. Gaffney, A. W. Kidder, and A. B. Ziffer. 1993. Florida reach ins: Environmental chambers for entomological research. Ame rican Entomologist Fall 1993: 177 182. Wells, J. D. 1991. Chrysomya megacephala (Diptera: Calliphoridae) has reached the continental United States: Review of its biology, pest status, and spread around the world. Journal of Medical Entomology 28: 471 4 73.

PAGE 95

95 Wells, J. D., and H. Kurahashi. 1994. Chrysomya megacephala (Fabricius) (Diptera: Calliphoridae) development: Rate, variation and the implications for forensic entomology. Japanese Journal of Sanitization and Zoology 45: 303 309. Wells, J. D., and L. R. LaMotte. 2001. Estimating the postmortem interval In J. H. Byrd and J. L. Castner [eds.], Forensic Entomology: The utility of arthropods in legal investigations. CRC Press, Boca Raton. Wickham, H. 2009. ggplot2: Elegant graphics for data anal ysis. Springer, New York. Zumpt, F. 1965. Myiasis in man and animals in the Old World: A textbook for physicians, veterinarians and zoologists. Butterworth, London. 267 pp.

PAGE 96

96 BIOGRAPHICAL SKETCH Susan Gruner was born in Wiesbaden, Germany in 1956 and graduated from high school in 1974. At the ripe old age of 40, she decided to return to college. She thought she would be a vet, but fell in love with insects. She received her B.S. degree at the Univ ersity of Florida in the last class of the second millennium. She was determined to find her way in the field of forensic entomology and in 2004 received her M.S. degree with funding from the NIJ. Being a glutton for punishment, Susan decided to get her do ctorate studying forensic entomology. The long, arduous task took many years and finally came to an end in 2014.