Rate of development of Hydrotaea rostrata under summer and winter (cyclic and constant) temperature regimes

Medical and Veterinary Entomology (2001) 15, 177±182 Rate of development of Hydrotaea rostrata under summer and winter (cyclic and constant) temperature regimes I. R. DADOUR *, D. F. COO K ² andn. WIRTH ³ * Zoology Department/Forensic Science Unit, University of Western Australia, ² Entomology Section, Agriculture Western Australia and ³ School of Biological Sciences, Curtin University, Western Australia Abstract. The black carrion y Hydrotaea rostrata Robineau-Desvoidy (Diptera: Muscidae) is a muscid, which occurs on approximately one-third of decomposing human corpses involved in homicide cases in south-western Australia. Work to date on its development rates is scant with only one published source available. The current study measured the precise developmental rates of this species using high repetition and sampling rates. A comparison was made of the developmental rates between constant and cyclic temperatures in winter and summer temperature regimes in south-western Australia. Developmental times for 90% of rst stage larvae to emerge as adult ies are: summer constant, 21.6 days, summer cyclic, 23.5 days, winter constant 64.5 days and winter cyclic, 48.3 days. These data will allow forensic entomologists to make more accurate determinations of post-mortem intervals in cases where H. rostrata life history stages are present. Key words. Hydrotaea rostrata, developmental rates, forensic entomology, postmortem intervals, Australia. Introduction Forensic entomology is the study of insects that are associated with a corpse, which are used to determine time since death in homicide cases and coroner's proceedings (Nuorteva et al., 1977; Anderson & van Laerhoven, 1995). Adult blow ies are usually the rst insects to arrive at a corpse, generally within several hours after death (Morris, 1994). Understanding the succession of adult ies visiting a corpse, the mode of reproduction of each y species (i.e. either larviposition or oviposition) and their subsequent development are paramount in obtaining an accurate measure of time since death. Hence, precise determination of developmental rates of dipterous insects has become an invaluable tool in crime ghting. A distinct correlation exists between the length of time that a corpse is exposed to insects and the number of waves of insects that may visit. Amongst the ies, there are usually three major waves arriving at the corpse, which either lay eggs (oviposition) or live larvae (larviposition). It has been suggested by Correspondence: Ian Dadour, Zoology Department, University of Western Australia, Nedlands, 6909, Western Australia. Fax: + 61 893801029; e-mail: idadour@cyllene.uwa.edu.au Monzu (1977) that in south-western Australia the primary wave of species includes Calliphora spp. and Lucilia spp. Chrysomya spp. and Sarcophagidae ( esh ies) are usually the next wave of blow ies to arrive and they are known as secondary species. The black carrion y Hydrotaea rostrata Robineau-Desvoidy is representative of the third wave of y species to arrive at a corpse, hence they are referred to as tertiary species. When estimating a post-mortem interval (PMI), two factors need to be considered. Firstly, the time it has taken for each y species to arrive at the corpse and secondly the rate at which the different y species develop through their respective life stages (Morris, 1994). The latter is in uenced predominantly by environmental factors (Howe, 1967), including diet, humidity, photoperiod, population density, rainfall and the presence of drug residues (West, 1951; Cohen, 1968; Beck, 1986; Hagstrum & Milliken, 1991; Goff et al., 1997). The two temperature measurements needed when dealing with the insects associated with a corpse, are the ambient temperature and the internal temperature of the corpse (West, 1951). Morris (1994) states that the higher the temperature in which the insect (at any developmental stage) is developing, the faster the rate of development of the insect. Typically, # 2001 Blackwell Science Ltd 177

178 I. R. Dadour et al. corpses are invaded by large masses of primary and secondary y larvae, which, during late second and third stage, generate enormous amounts of metabolic heat and so the temperature at which they are developing is usually higher than the ambient temperature (Goff, 1993; Morris, 1994). The internal temperature of a corpse compared to the outside temperature will also be different depending on the stage of decomposition (O'Flynn, 1983) Diurnal shifts in temperature may also retard or accelerate larval growth when compared to that at a constant temperature, depending upon the target y species (Gullan & Cranston, 1994), even when the mean of the cyclic temperatures equals the constant temperature (O'Flynn, 1983; Byrd & Butler, 1996, 1997, 1998; Erzinclioglu, 1996). One y species of forensic importance in Australia is H. rostrata, which lays eggs onto corpses during the late stages of decomposition, but does not form large maggot masses (O'Flynn, 1983). A number of homicide cases involve the discovery of bodies several weeks to several months after death and H. rostrata is invariably the most important y species in determining a PMI. There is little information available on the developmental rates for H. rostrata, hence the aim of this study was to compare the rates of development of all the life cycle stages of H. rostrata, under both constant and cyclic temperatures, typical of summer and winter conditions in south-western Australia. Methods Adult H. rostrata were collected from the eld (Harry Warring Marsupial Reserve, Jandakot between March and April 1998) and reared in the laboratory. Adult ies were kept in cloth-covered cages (40 3 20 3 20 cm) and supplied with sugar and water. The cages were kept in a constant temperature room at 26/22 C: LD 12:12 h. Every 7d a 50 : 50 mixture of minced ox liver and jellymeat (media) was used to feed and induce oviposition by the female ies and support larval development. The adult ies used for this experiment were F 3 generation. Design Egg and larval development rates of H. rostrata were measured inside constant temperature rooms under exposure to four temperature regimes. These temperatures represented winter constant (WC) (15 C), summer constant (SC) (25 C), winter cyclic (WF) (18/12 C) and summer cyclic (SF) (30/ 19 C) temperature regimes. The cyclic temperatures were calculated using the daily temperature averages (minimum and maximum) and the constant temperatures were the averages of the cyclic temperatures. The Bureau of Meteorology (1999) provided the summer regimes data from December, January and February and the winter regimes data from June, July and August. Temperature changes inside each constant temperature room took 30 min to equilibrate. The photoperiod was set at LD 12 : 12 h. Egg hatch Three containers of media (as described above) were placed into each cage of 100 adult ies (mixed sex). These ies were maintained at 25 C with continuous light. The media were checked every 2 h and as soon as eggs were found they were placed on moist lter paper in Petri dishes. Ten replicates of 100 eggs per replicate were placed in the SC and WC temperature regimes and 20 replicates in the SF and WF temperature regimes. In the cyclic temperatures, 10 replicates were exposed to the minimum and 10 replicates to the maximum temperatures for the rst 12 h. In the summer temperature regimes the egg hatch was recorded every 2 h until 90% of the eggs from each replicate had hatched. For the winter temperature regimes the eggs were checked every 12 h until 48 h had passed and then every 2 h until 90% of the eggs from each replicate had hatched. Larval development Eggs deposited over a 19-h period were collected from three cages (as described above) containing 100 H. rostrata adults. Each cage had a 50 : 50 sex ratio of male and female ies. Thirty eggs were individually placed onto a 1.5 cm square of black lter paper, with a ne paint brush. The lter paper was placed with the eggs facing down onto the different amounts of medium described below. A total of 225 replicates were set up for each temperature regime. In the cyclic temperature regimes the eggs were exposed to the maximum temperature for the rst 12 h. Sampling regimes were worked out to best suit each temperature regime and the approximate duration of larval development for the summer and winter temperature regimes expected. These were 20 and 60 days, respectively, based on O'Flynn (1983). For the rst 8 days of the winter temperature regimes and the rst 4 days of the summer temperature regimes, vial lids (2.5 cm in diameter) with 2 g of media, were used. This expedited the recovery of the newly hatched larvae. Petri dishes (8.5 cm in diameter) with 30 g of media were used for the rest of the experiment. The vial lids and Petri dishes were placed into large boxes (35 3 35 3 5 cm) with plastic lids. The undersides of the lids were sprayed with water to maintain high humidity until the eggs had hatched. Once rst stage larvae were present, the plastic lids were removed and replaced with ne mesh lids. Each replicate was sprayed daily with water to prevent desiccation. When the larvae had moulted into third stage larvae, the Petri dishes were placed into individual containers with yellow sand in the bottom, to allow the larvae to leave the food source and commence pupation. The containers were covered with a mesh lid to trap the adult ies following emergence. Sampling When sampling the rst 8 days of the winter regimes and the rst 4 days of the summer regimes, the entire contents of the vial lids (media and larvae) were placed in boiling water,

Black carrion y development 179 the larvae removed and placed in 70% alcohol. Larvae in the Petri dishes (which were second and third stages) were physically removed from the media and placed into boiling water and then into 70% alcohol. Following pupation, half of the pupae collected at each sampling point were pricked with a ne pin and preserved in 70% ethanol. The other half was placed back into the sand to allow the adult ies to emerge. Larval measurements All larvae that were collected and preserved at each sampling point were examined under a dissecting microscope to record the stage. In addition, ve larvae from each replicate were measured under the microscope using an ocular micrometer, accurate to 0.5 mm. Table 1. Time taken for 90% of eggs to hatch under constant and cyclic temperature regimes. The temperatures listed under the cyclic regime are those in which the eggs were exposed to for the rst 12 h. Regime Temperature ( C) Hatch (mean 6 SE) (h) SC 25 26.0 6 0.3 SF 30 19.5 6 0.1 SF 19 25.0 6 0.2 WC 15 56.2 6 0.8 WF 18 48.9 6 0.8 WF 12 56.3 6 0.8 for the SC regime. The time required for 90% of third stage larvae to puparia for the SF regime was signi cantly longer than the SC regime (t 1503 = 3.4, P < 0.01; Table 2). Statistics The data on developmental rates (cumulative), and larval length were analysed using analysis of variance and t-test. Normality of residuals resulting from analysis of variance was checked using normal probability plots (Zar, 1984). All analyses were conducted using Genstat (1997). Results Egg hatch The duration of the egg stage was de ned as being from when eggs were laid until 90% of the eggs had hatched. There was a signi cant difference when comparing the SF regime when exposed to 30 C for the rst 12 h with the SC regime (t 35 = 11.4, P < 0.01), and the WF regime when exposed to 18 C for the rst 12 h with the WC regime (t 83 = 5.4, P < 0.001; Table 1). Larval development: summer regimes The larval development rates for the summer regimes were also signi cantly different. There were marked differences in the duration of each individual larval stage. The moult from rst to second stage for the SF regime began at 2.3 days (56 h) and was signi cantly different compared with the SC regime, which began at 3.7 days (88 h). The time required for up to 90% of larvae to moult from rst to second stage for the SC regime was signi cantly longer than the SF regime (t 69 = 16.6, P < 0.001). At the second to third larval moult, although the SF regime began at 4 days (96 h) and the SC regime began at 3 days (128 h), there was a signi cantly longer mean time for 90% to moult for the SF regime when compared with the SC regime (t 49 = 7.0, P < 0.001). Again, the trend remained the same during development from third larval stage to puparia. It began at day 11.4 (273 h) for the SF regime and day 10 (240 h) Larval development: winter regimes The larval development rates for the winter experiments differed signi cantly. Under the WF regime it took 5.6 days (134 h) for larvae to begin moulting into second stage larvae and up to 8.6 days (206 h) for larvae in the WC regime. The mean time for 90% of larvae to reach second stage in the WF regime was signi cantly faster than larvae in the WC regime (t 52 = 22.9, P < 0.001). The second stage larvae began to moult at 10.6 days (254 h) in the WF regime and 13.6 days (326 h) in the WC regime. The mean time for 90% of the larvae to reach third stage in the WF regime was again signi cantly faster than larvae in the WC regime (t 79 = 25.8, P < 0.001). The third stage was the longest stage of larval development, commencing 24.6 days (590 h) after oviposition in the WF regime and 38.6 days (926 h) in the WC regime. In the WF regime the mean time for 90% of larvae to complete pupariation was also signi cantly faster when compared with the WC regime (t 325 = 24.2, P < 0.001; Table 2). Total pre-adult development time Adult ies began emerging from the SF regime on day 21 (504 h) following egg hatch compared with day 19 (456 h) in the SC regime. Under the SF regime, mean time for 90% of adult ies to emerge was signi cantly longer when compared with the SC regime (t 1071 = 46.7, P < 0.001). Adult ies began emerging from the WF regime at day 44 (1056 h) compared with day 60 (1440 h) for the WC regime. Mean time to emergence (i.e. the period from when 90% of adult ies had emerged from the pupae) was signi cantly faster in the WF regime when compared with the WC regime (t 75 = 46.7, P < 0.001). Larval growth rates Table 3 shows the mean larval sizes for all three larval stages of H. rostrata in each of the four temperature regimes

180 I. R. Dadour et al. Table 2. Developmental time (h) (cumulative time for 90% of each stage to change) from egg hatch to emergence (mean 6 SE). Figures in parentheses show developmental time in days. Regime 1st±2nd stage larvae 2nd±3rd stage larvae 3rd stage larvae ± pupation Total time to emergence SC 95.5 6 1.1 (3.9) 128 (5.3) 372.8 6 2.5 (15.5) 519.5 6 1.1 (21.6) SF 69.3 6 1.1 (2.9) 132.0 6 0.6 (5.5) 383.5 6 1.9 (16) 563.1 6 1.2 (23.5) WC 221.7 6 2.2 (9.2) 392.4 6 3.2 (16.3) 1177.3 6 9.9 (49.1) 1548.6 6 7.0 (64.5) WF 148.8 6 2.3 (6.2) 285.8 6 2.6 (11.9) 820.8 6 10.9 (34.2) 1159.9 6 4.4 (48.3) Table 3. A summary of the length (mm) and statistical results for each stage of H. rostrata larvae reared over four different temperature regimes. 1st stage larva 2nd stage larva 3rd stage larva Temperature regime Range Mean 6 SE Statistical test Range Mean 6 SE Statistical test Range Mean 6 SE Statistical test WC (15 C) 1.2±3 2.25 6 0.15 t = 0.527 4±8.3 6.73 6 0.26 t = 0.699 11.8±16.4 14.9 6 0.26 t = 0.836 n = 87 P > 0.05 n = 114 P > 0.05 n = 303 P > 0.05 WF (12±18 C) 1.3±3 2.26 6 0.17 4.4±8.6 6.35 6 0.33 10.6±16.8 15.5 6 0.3 n = 72 n = 72 n = 158 SC (25 C) 1.3±2.8 2.37 6 0.15 t = 0.531 4.4±7.7 5.47 6 0.27 t = 1.943 13.5±15.8 14.8 6 0.23 t = 0.604 n = 118 P > 0.05 n = 64 P > 0.05 n = 566 P > 0.05 SF (19±30 C) 1.6±3 2.23 6 0.15 5.5±8.3 7 6 0.25 11.5±15.7 14.4 6 0.25 n = 79 n = 173 n = 616 and shows that maximum size is not in uenced by temperature. Each stage has similar lengths across all four temperature regimes. When comparing the summer and winter temperature regimes, the mean sizes of each larval stage were not signi cantly different (Table 3). Discussion Predictably, there was great variation in the development rates of H. rostrata exposed to either summer or winter temperature regimes. More importantly, however, this study demonstrated differences in the development rates between constant and cyclic temperature regimes within each season. If we compare the total development times then this is particularly evident for the winter regimes. In a WC regime (15 C) the mean time for development from egg to adult took approximately 64.5 days, whereas under a WF regime (12±18 C) it only took approximately 48.3 days. The only other study on this species gave a large developmental range (49±70 days) at a constant temperature of 15 C (O'Flynn, 1983). In addition, under summer temperature regimes, the mean time for development was also signi cantly different between constant and cyclic regimes but only by 1.5 days. Under the SC regime (25 C) the development from egg to adult took approximately 21.6 days, whereas under the SF regime (19± 30 C) it took approximately 23.5 days. Again, the only contrasting study on this species by O'Flynn (1983) recorded the length of development from egg to adult under summer temperature regimes was 28 and 18 days at constant temperatures of 20 C and 28 C, respectively. There were some key differences between the ndings of this study and O'Flynn (1983). For example, the WF regime took up to 56 h for 90% of the eggs to hatch and approximately 49 h under the WC regime, a difference of approximately 7 h. By comparison, in O'Flynn (1983), at a constant temperature of 15 C the egg stage took 72 h. When comparing the summer temperature regimes the trend was reversed, with egg hatch under the SF regime being approximately 6.5 h faster. Eggs exposed to the cyclic summer temperature regime (SF) would have experienced a high maximum temperature for longer (30 C for rst 12 h followed by 19 C for 7.5 h) than eggs under the SC regime. In contrast, the small time difference (1 h) between the SF and the SC regime may be due to the eggs being exposed to cooler temperatures for a longer period than the maximum temperature (19 C for 14 h). Similarly, under the WF regime, eggs would take longer to develop when exposed to 12 C for the rst 12 h period (12 C for 32 h) when compared with eggs exposed to 18 C for the rst 12 h (12 C for 24 h). The duration of each larval stage varied between all four temperature regimes, especially in the WC regimes. The difference in developmental time increased between WF and WC at each moult. From rst to second stage larvae the difference was 3 days, by the third stage it was 4.4 days, and from oviposition to puparia it had increased to 15 days. Larvae developing under the SF regime had the shortest rst stage, differing from SC by only 24 h. This difference in larval development decreased during the second to third stage (5 h)

Black carrion y development 181 and the third stage to pupal instar (12 h) but there was 48 h difference during the pupal instar. Comparing the constant developmental times with O'Flynn (1983) revealed large differences. At a constant temperature of 15 C, O'Flynn (1983) recorded that the rst and second larval stage of H. rostrata took about 3 and 4 days, respectively. These results are in contrast to the present study, in which it took approximately 9.2 days for the rst stage and 16.3 days for the second stage to complete development. The total larval development for the WC regime (15 C) was 49.1 days, whereas O'Flynn (1983) recorded that it took 25 to > 49 days to develop through all larval stages. Time to complete development to the third stage at the SC temperature regime took 15.5 days. O'Flynn (1983) did not record individual larval stages at higher temperatures, but at constant temperatures of 20 C and 28 C he recorded that it took 13±16 days and 9±11 days, respectively, to complete the larval development. Based on O'Flynn (1983), one would predict it would take between 10 and 13 days to complete larval development at 25 C constant temperature. From previous studies of ies comparing development under cyclic and constant temperature regimes, the effects on development rate are species speci c (Byrd & Butler, 1996, 1997, 1998). In some species development is accelerated, in other species it is retarded, and there can sometimes be no measurable difference. Davies & Ratcliffe (1994) observed the development of the larval stages of ve blow y species of forensic importance at both cyclic and constant temperature regimes. They found that development was accelerated in three species (Calliphora vomitoria, Phormia terraenova and Lucilia sericata) at the cyclic temperature regime when compared to a constant temperature. However, larval development was retarded in C. vicina under cyclic temperature regimes when compared to the respective constant temperature. Greenberg (1991) found that development from egg to adult was retarded in L. sericata, and some other blow y species, when reared at cyclic temperatures compared to the respective constant temperature. The studies comparing development rates at cyclic vs. constant temperatures subjected y life history stages to maximum and minimum temperatures closely associated with normal diurnal temperature uctuations rather than one continuous temperature. Developmental rates under constant temperatures are still very important from a forensic perspective especially when corpses are located in dwellings. The duration of pupal stage in H. rostrata was also in uenced by temperature. The length of the pupal stage for the WC regime was approximately 15.4 days, only one day longer than the WF regime (»14 days). The duration of the pupal stage for the summer regimes also only differed by approximately one day (SC, 6.1 days vs. SF, 7.5 days). Dallwitz (1984) is the only published work on pupal development of a blow y under constant and cyclic temperature regimes. He found that L. cuprina pupae reared under cyclic temperatures and respective constant temperatures up to 30 C showed no difference in developmental time up to 30 C. However, at temperatures > 30 C, pupae reared under cyclic temperatures developed at a faster rate than the corresponding constant temperature. The data produced from these experiments will assist forensic entomologists to make more accurate determinations of PMIs in cases where H. rostrata life history stages are present. Both an absolute estimation of the duration of each stage and calibrated larval length of the H. rostrata life cycle will be useful in estimating PMIs. Acknowledgements Thanks to all the technical staff who helped out with the logistics of the work, Joselyn Fissioli, David Cousins, Jeremy Lindsey, Wayne Morris, and Bengt Thaung. The Australian Research Council (C19941033) and the National Institute of Forensic Science funded this research. Facilities were supplied by Agriculture Western Australia and University of Western Australia. This paper formed part of a 3rd year project at Curtin University of Technology. References Anderson, G.S. & van Laerhoven, S.L. (1995) Initial studies on insect succession on carrion in southwestern British Columbia. Journal of Forensic Science, 41, 617±625. Beck, S.D. (1986) Insect Photoperiodism. 2nd edn. Academic Press, New York. Byrd, J.H. & Butler, J.F. (1996) Effects of temperature on Cochliomyia macellaria (Diptera: Calliphoridae) development. Journal of Medical Entomology, 33, 901±905. Byrd, J.H. & Butler, J.F. (1997) Effects of temperature on Chrysomya ru facies (Diptera: Calliphoridae) development. Journal of Medical Entomology, 34, 353±358. Byrd, J.H. & Butler, J.F. (1998) Effects of temperature on Sarcophaga haemorrhoidalis (Diptera: Sarcophagidae) development. Journal of Medical Entomology, 35, 694±698. Cohen, B. (1968) The in uence of larval crowding on the development time populations of Drosophila melanogaster on chemically de ned medium. Canadian Journal of Zoology, 46, 493±497. Dallwitz, R. (1984) The in uence on constant and uctuating temperatures on development rate survival of pupae on the Australian sheep blow y Lucilia cuprina. Entomologica Experimentia Applicata, 36, 89±95. Davies, L. & Ratcliffe, G.G. (1994) Development rate of some preadult stages in blow ies with reference to low temperatures. Medical and Veterinary Entomology, 8, 245±254. Erzinclioglu, Z.Y. (1996) Blow ies. The Richmond Publishing Company Ltd, Slough, U.K. Genstat, 5 (1997) Genstat 5. Release 4.1. Lawes Agricultural Trust, Rothamsted Experimental Station, U.K Goff, M.L. (1993) Estimation of postmortem interval using arthropod development and successional patterns. Forensic Science Review, 5, 81±94. Goff, M.L., Miller, M.L., Paulson, J.D., Lord, W.D., Richards, E. & Omori, A.I. (1997) Effects of 3,4-Methylenedioxymethamphetamine in decomposing tissue in the development of Parasarcophaga ru cornis (Diptera: Sarcophagidae) and detection of the drug in postmortem blood, liver tissue, larvae and puparia. Journal of Forensic Science, 42, 276±280.

182 I. R. Dadour et al. Greenberg, B. (1991) Flies as forensic indicators. Journal of Medical Entomology, 28, 565±577. Gullan, P.J. & Cranston, P.S. (1994) The Insects. An Outline of Entomology. Chapman & Hall, London. Hagstrum, D.W. & Milliken, G.A. (1991) Modeling differences in insect development times between constant and uctuating temperatures. Annals of the Entomological Society of America, 84, 369±379. Howe, R.W. (1967) Temperature effects on embryonic development in insects. Annual Review of Entomology, 12, 15±42. Monzu, N. (1977) Coexistence of Carrion Breeding Calliphoridae (Diptera) in Western Australia. PhD Thesis, University of Western Australia, Perth. Morris, B. (1994) Forensic entomology. Expert Evidence, 91A, 8-5291±8-5381 (ed. by I.R. Freckleton and H. Selby). The Law Book Company Ltd, Sydney. Nuorteva, P. (1977) Sarcosaprophagous Insects as Forensic Indicators, Vol. II. Forensic Medicine: a Study in Trauma and Environmental Hazards (ed. by C. G. Tedeschi, W. G. Eckert and L. G. Tedeschi), pp. 1072±1095. Saunders, Philadelphia. O'Flynn, M.A. (1983) The succession and rate of development of blow ies in carrion in Southern Queensland and the application of these data to forensic entomology. Journal of the Australian Entomological Society, 22, 137±148. West, L.S. (1951) The House y: its Natural History. Medical Importance and Control. Comstock Publishing Company Inc, New York. Zar, J.H. (1984) Biostatistical Analysis. 2nd edn. Prentice Hall, Englewood Cliffs, NJ. Accepted 16 February 2001