Medical and Veterinary (2006) 20, 106–114 doi: 10.1111/j.1365-2915.2006.00600.x

Larval growth rates of the blowfly, Calliphora vicina, over a range of temperatures

S. E. DONOVAN1 ,M.J.R.HALL1 ,B.D.TURNER2 and C . B . MO NC R I E F F 1 1The Natural History Museum, London, 2Department of Life Sciences and Forensic Science Unit, Kings College, University of London, U.K.

Abstract. Blowfly larvae (Diptera: ) fulfil an important ecological function in the decomposition of remains. They are also used extensively in , predominantly to establish a minimum time since death, or a minimum post-mortem interval, using the larval length as a ‘biological clock’. This study examined the larval growth rate of a forensically important , Calliphora vicina Robineau-Desvoidy (Diptera: Calliphoridae) at tem- peratures of between 4Cand30C, under controlled laboratory conditions. The laboratory had been trapped initially in London, U.K. The minimum devel- opmental temperature was estimated to be 1C and 4700 accumulated degree hours (ADH) were required for development from egg hatch to the point of pupariation. Lines fitted to the laboratory larval growth data were found to adequately explain the growth of larvae in the field. The nature of variation in growth rates from geographically isolated populations is discussed. Key words. Calliphora vicina, Calliphoridae, forensic entomology, larval growth, temperature, U.K.

Introduction Many species of blowfly are attracted to dead bodies of Forensic entomology involves the interpretation of and humans (Greenberg & Kunich, 2002). If evidence in legal investigations. It has many applications allowed access to a body, the adults will feed on any secre- (Byrd & Castner, 2001), but the most important and sig- tions, including blood, and gravid females will rapidly lay their eggs on the corpse. In the case of Calliphora vicina nificant is in investigations involving human corpses. recovered from a body, predominantly fly larvae (Robineau-Desvoidy), at temperatures above 15 C the eggs and beetles, can provide information on the conditions hatch after about 24 h, whereupon the larvae begin to feed experienced by the body following death, and sometimes, on the body tissues. Once fully grown, the post-feeding in cases of neglect, prior to death. Most commonly though, larvae usually migrate away from the body to pupariate, fly larvae are used to estimate the minimum time since although some species will pupariate on or in the immediate death, that is the minimum post-mortem interval (PMI). vicinity of the body (Greenberg, 1991). At any of these This can be inferred from the species composition of insects stages the police, or forensic entomologists, can remove found on the body, as the types of insects found on and in a insect specimens, some of which will be killed and the rest body follow a fairly predictable succession, first reported by reared to the adult stage to facilitate species identification. Me´ gnin (1894). However, when the first generation of off- The size of a , usually represented by its length, is spring from the first flies to locate the corpse are still on or related to its age as a function of time and temperature, around it, a more precise figure can be obtained by estimat- and so should be able to provide a minimum age for the ing the age of the largest, and therefore oldest, larvae on the oldest (longest) larvae and therefore a minimum PMI. body. Their greatest ages indicate the time when flies first Different species of blowfly grow at different rates, and laid their eggs on the corpse. considerable effort has gone into generating developmental

Correspondence: Dr Sarah Donovan, School of Biological Sciences, University of Plymouth, Drakes Circus, Plymouth, Devon PL4 8AA, UK. Tel.: þ 44 (0)1626 325880; fax: þ 44 (0)1626 325616; e-mail: [email protected]

106 # 2006 The Authors Journal compilation # 2006 The Royal Entomological Society Larval growth in Calliphora vicina 107 data for a number of species of forensic importance: Calliphora growth with those of larvae from field populations and also alpina Zett. (Davies & Ratcliffe, 1994); Calliphora vicina to compare the results with published data for this species, (Kamal, 1958; Reiter, 1984; Greenberg, 1991; Davies & to consider the potential importance of geographical varia- Ratcliffe, 1994); (Linnaeus) (Kamal, tion on development rate. 1958; Greenberg & Tantawi, 1993; Davies & Ratcliffe, 1994); Chrysomya megacephala (Fabricius) (Wells & Kurahashi, 1994); Chrysomya albiceps (Wiedemann) (Grassberger et al., Methods 2003); Chrysomya rufifacies (Macquart) (Greenberg, 1991; Byrd & Butler, 1997); Cochliomyia macellaria (Fabricius) Source of larvae (Greenberg, 1991; Wells & LaMotte, 1995; Byrd & Butler, 1996); Lucilia sericata Meigen (Kamal, 1958; Ash & Colonies of C. vicina were used for the studies. These had Greenberg, 1975; Greenberg, 1991; Wall et al., 1992; Davies been collected initially from two locations in London, & Ratcliffe, 1994; Grassberger & Reiter, 2001); U.K.: the wildlife gardens of the Natural History (Meigen) (Kamal, 1958; Greenberg, 1991; Byrd & Allen, 2001); Museum (NHM), South Kensington and from the roof of terraenovae (Robineau-Desvoidy) (Kamal, South Bank University (SBU), near Waterloo Station. 1958; Greenberg, 1991; Greenberg & Tantawi, 1993; Davies Adults were kept at 18 4C and LD 12 : 12 h. Newly & Ratcliffe, 1994; Grassberger & Reiter, 2002b). Similar infor- emerged adults were provided with protein (fresh lamb liver mation has recently been published on two species of fleshfly of exudate) to allow egg maturation. Flies had access to water forensic importance (Diptera: Sarcophagidae) (Byrd & Butler, and granulated sugar cubes at all times. Lamb liver was 1998; Grassberger & Reiter, 2002a). added to the cages about a week after adult emergence, to Development is usually quantified as accumulated degree allow oviposition. After about 12 h the eggs were removed, hours (ADH) or days (ADD) above some minimum tem- rinsed in distilled water and then placed on damp filter perature below which development ceases (Sharpe & paper in Petri dishes. First-stage larvae were used within DeMichele, 1977). However, for C. vicina this minimum is 3 h of hatching (90% within 2 h). not universally agreed and published values range from 6C (Haskell et al., 2000) down to 2C (Vinogradova & Marchenko, 1984; Greenberg, 1991). Davies and Ratcliffe Measurement of growth rates observed eggs of C. vicina to hatch at 3.5C and therefore concluded that the minimum threshold must be below Using a fine, dampened paintbrush, groups of 20 first- 3.5C (Davies & Ratcliffe, 1994). Ames and Turner studied stage larvae were transferred into 20 mL polystyrene cups, development of C. vicina at temperatures down to 1C, but each containing 20 g of roughly chopped lamb liver, which concluded that there was no clear cut minimum develop- was dampened periodically to prevent desiccation. Each ment threshold (Ames & Turner, 2003). cup was covered with a fine, nylon mesh, held in place An alternative approach to estimating the age of larvae is with an elastic band. Six cups were placed in a larger plastic the use of isomegalen diagrams (Reiter, 1984; Grassberger & box (25 15 10 cm deep) with a loose lid. The box Reiter, 2001). From these diagrams it is possible to estimate contained a layer of water to maintain a sufficiently high a larva’s age from its length, assuming that it has developed humidity around the larvae; the average relative humidity throughout at a constant temperature. In reality, tempera- was above 40% in all cases. Groups of eight plastic boxes at tures will have varied, and dealing with fluctuating tempera- a time were placed in total darkness in incubators set at a tures can be problematic because an average temperature has specified constant temperature. to be calculated for the entire period of development. This Incubator temperatures encompassed the range likely to problem can be overcome to some extent by using accumu- be experienced by larvae in the U.K.: 4, 10, 14, 19, 23, 26 and lated degree days or degree hours, despite the potential 30C (each 1.5C). In a pilot study, all larvae reared at uncertainty if meteorological data is of poor quality. 35C had died before pupariation. The boxes were venti- A further problem in the use of knowledge of develop- lated daily, and moved within the incubators to minimize ment rates to estimate larval age is that populations of the the effect of any systematic temperature gradients. A data- same species can differ physiologically depending on their logger [Tinytag Plus, Gemini Data Loggers Ltd (Chichester, geographical origin. For example, populations of C. vicina U.K.)] was placed in each incubator to record temperature from the south of England and those from Finland differ and relative humidity at 10-min intervals. significantly in their response to photoperiod and To validate the laboratory results, field trials were run in temperature. There is strong evidence that the differences August, October and November 2001. For these, individual are genetic (McWatters & Saunders, 1996, 1998). There cups of larvae were prepared as described above. These appears to be considerable variation in the accumulated were placed in a wire cage under shrubs in the grounds of degree hours required for development by geographically the Wildlife Garden at The Natural History Museum. The different populations of forensically important fly species cage provided protection from vertebrate scavengers, and (Higley & Haskell, 2000). The aim of this study, therefore, was also covered with nylon mesh to prevent local wild was to describe larval growth of laboratory populations of blowflies from laying eggs on the liver. As before, a data- C. vicina derived from London, to compare these rates of logger recorded temperature and relative humidity.

# 2006 The Authors Journal compilation # 2006 The Royal Entomological Society, Medical and Veterinary Entomology, 20, 106–114 108 S. E. Donovan et al.

Sampling against temperature back to the point where the rate of development reaches zero can underestimate the develop- The larvae were sampled throughout their development at ment threshold because the curve becomes non-linear at specified intervals (Table 1). On each sampling occasion, at lower temperatures. Here, the developmental threshold least three or four cups of 20 larvae were removed (as was determined by fitting linear and second and third- Ò numbers allowed). All the larvae were removed from the order polynomial regressions (UNISTAT 5.0 statistical soft- liver, rinsed in water at room temperature and then killed ware, 2000) to a plot of recorded accumulated degree hours by immersion in boiling water for 30 s. They were then against the mean of loge larval length transformed back to preserved in 80% ethanol, and their lengths measured within original scale, measured at each of the seven constant tem- 1 h using an eyepiece micrometer on a light microscope. peratures used. The regression lines were fitted using a Larval lengths change by only about 1.2% within the first range of theoretical minimum developmental temperatures, hour of preservation in 80% ethanol (Adams & Hall, 2003). from 2Ctoþ4C and maximum R2 was used as the In the few instances where immediate measurement proved criterion to calculate best fit. The best-fit curves were impractical, the preserved larvae were stored at 20Cuntil third order polynomials, obtained when using theoretical they could be measured. Earlier trials had shown that this minimum developmental thresholds of 1.0C and 1.5C, does not affect size. The time required for eggs to hatch was which generated R2 values of, respectively, 0.891 and not recorded, but larvae were used soon after emergence. 0.897. These curves indicated that the developmental mini- To allow pupariation, between three and six cups of larvae mum lay between 1.0C and 1.5C but, for simplicity, a had their covers removed and were placed in individual sealed value of 1C was used in subsequent analysis. but ventilated containers with sand at the bottom. Post- To quantify the change in larval length over time, for each feeding larvae migrated and pupariated in the sand. Once batch of larvae, the mean of loge length for all larvae and for they had pupariated, or died, they were transferred to a poly- the larvae with lengths above the lower quartile, were deter- styrene cup, with a covering of barely damp sand, again with mined. A range of non-linear models were then fitted to the a nylon mesh cover. The adult flies were killed and sexed. data by least squares regression (GenStat 8.1, 2005). This was Their size was recorded by measuring the cm-du cross-vein undertaken both for all the data and for data up to 2400 length (Smith & Wall, 1997) using an eyepiece micrometer. accumulated degree hours only, as the latter equates to the This measurement was also taken from wild caught individ- feeding phase. Using the line fitted to the laboratory data, uals from the same locality (NHM wildlife garden) to estimated accumulated degree hours were then calculated for compare the sizes of laboratory and field-derived flies. the larvae reared outside based on their lengths. The observed and estimated accumulated degree hours were then compared.

Statistical analyses Results A mean value of log larval length was obtained for the e Size and sex ratio of experimental population larvae from each cup and, from these, for the three or four cups removed at the same time, as they could not be Adults from both the laboratory and field growth experi- assumed to be independent samples. This was because the ments were not significantly different in size from wild flies cups of larvae had been set up at a small number of fixed caught in the NHM wildlife garden, as determined from the times from a common pool of larvae and shared the same length of the cm-du cross vein (F ¼ 1.22, P > 0.1 and conditions. Groups of cups removed at different times were F ¼ 1.21, P > 0.1, respectively). In addition, the male : considered to be independent samples. female sex ratio of the adult flies produced was not signifi- In estimating the developmental minimum temperature, cantly different from 1 : 1 (w2 ¼ 1.71, P > 0.1). There were linear extrapolation of a plot of daily developmental rate no significant, systematic differences in growth rates between larvae from different sources (SBU or NHM) or generations. Table 1. Sampling times for larvae of Calliphoora vicina reared at each temperature Larval growth Ambient temperature Sampling (C) times (h) Larval growth showed a pattern of rapid increase in length, followed by shrinking prior to pupariation 4 0.5, 48, 120, 192, 288, 360, 452, 528, 623, 720, 864, 1056 (Fig. 1). A reasonable fit to this pattern was obtained 10 0.5, 29, 67, 117, 170, 216, 288, 338, 361, 432, 456, 600 using the equation: 14 0.5, 48, 67, 78, 101, 117, 148, 195, 286, 315 ð Þ¼ þ ð d = 19 0.5, 17, 25, 40, 49, 89, 113, 120, 140, 166, 191, 215 loge length a bADH 1 ADH 23 0.5, 24, 48, 72, 96, 120, 168, 192, 216, 223 ð1 þ 0:25bÞÞ=ð1þbdADHÞ; ð1Þ 26 0.5, 25, 44, 53, 68, 77, 97, 116 ¼ 30 0.5, 17, 25, 32, 43, 52, 67, 74, 95, 113, 139 where ADH accumulated degree hours and values for a, b, d and b are given in Table 2. This fitted the growth at

# 2006 The Authors Journal compilation # 2006 The Royal Entomological Society, Medical and Veterinary Entomology, 20, 106–114 Larval growth in Calliphora vicina 109

25

20

15

10 Geometric mean length (mm)

Fig. 1. The geometric mean length of lar- vae of the blowfly Calliphora vicina against 5 accumulated degree hours (ADH) to a maximum 5400 ADH. Each data point represents the mean length for a batch of (20) larvae reared in a single cup for a range of temperatures. The fitted hyper- bolic function (solid line) with 95% con- fidence limits of the model (dashed lines) 600 1200 1800 2400 3000 3600 4200 4800 5400 and for the observations (dotted lines) are shown (r2 ¼ 0.943). Observed ADH each of the seven temperatures considered, providing evi- Discussion dence for a common underlying growth mechanism at work at all temperatures (Figs 2a–g). The influence of slowly The results reported here gave a maximum accumulated developing laggards was excluded in further analyses by degree hours of 4700 for larval development of C. vicina, considering only the larvae with lengths above the lowest from newly hatched first stage larvae to the point of pupar- quartile (Fig. 3). iation, assuming a base temperature of 1C. This is close to When the data from the feeding phase only is used other ADH values for the same developmental interval (Fig. 4), it is possible to read off the accumulated degree reported in the literature. Hence, reworking the data pre- hours predicted to be required for a larva to reach a given sented by Greenberg (1991; Table 3) with a base tempera- size and therefore to determine the age of the larva, as long ture of 1C gave ADHs of 4104–4440 in the range of 12.5C as the temperatures experienced by that larva are known. to 25C. Similarly, reworking the data of Kamal (1958; The line fitted to the laboratory data also fitted the field data Table 1) with a base temperature of 1C gave an ADH of well and there was a good relationship between the ADH 5654 at 26.7C. predicted from the regression line and the actual ADH Using ADH gives an immediate advantage over the type calculated from the field data for the feeding phase of model presented in the isomegalendiagram (see Reiter, (Fig. 5), with departure from the 45 degree line being incon- 1984), as it can readily be used in natural situations where sequential (F ¼ 2.12, d.f. ¼ 2,11, P ¼ 17%, r ¼ 0.185). the temperature fluctuates. In addition, 95% confidence For four samples, the observed mean larval size exceeded limits were generated, based on the data, indicating where the maximum value predicted by the size : ADH curve (Fig. 4) the mean larval length of a cohort of individuals would lie. and therefore a predicted ADH is not provided by the curve. This enables limits to be set around estimates of larval age, However, in practice a predicted ADH can be estimated by which is of particular importance in a forensic context. To reducing the mean size to the ‘maximum value’ predicted by establish a PMI it is important to measure the largest – and the size : ADH curve (2354 ADH, fig. 4). This approximation therefore oldest – individuals, so the data of greatest foren- was applied to the four outlier samples (Fig. 5). sic value are based on the larger individuals (in this case the

# 2006 The Authors Journal compilation # 2006 The Royal Entomological Society, Medical and Veterinary Entomology, 20, 106–114 110 S. E. Donovan et al.

Table 2. Calculated values for the parameters in equation 1 for each of the four sets of larval growth data

Larval population data ab db

All data, max 5400 ADH (Fig. 1) 0.6320 0.002042 0.00016663 1.797 Upper three quartiles of data, max 5400 ADH (Fig. 3) 0.6580 0.002098 0.00016868 1.849 Upper three quartiles of data, max 2400 ADH (Fig. 4) 0.7125 0.001516 0.0002531 0.984

ADH, accumulated degree hours. largest 75%; Figs 3 and 4). There is inherent variability in better fit to the data, the single one that described larval growth rates, as described by Davies & Ratcliffe (1994), growth over the entire range of temperatures was of more with a small proportion of the individuals growing at a practical benefit. However, the potential for variation of much slower rate than the rest. ADH with temperature requires further study, because over There was some evidence that at temperatures above longer developmental periods than were measured in detail 20C maggot growth was underestimated, whereas at tem- here (egg to adult) there is evidence that ADH can vary, peratures below 20C it was overestimated (Fig. 2). being greater at lower temperatures (Greenberg, 1991). Therefore, other lines, not detailed here, were fitted to the In terms of estimating larval age, there are potential combined data of temperatures above 20C and those problems in confusing the two phases – feeding and post- below 20C. Although these both showed a marginally feeding – above a certain length. However, it is possible to

(g) 30° C 20 15 10 5

24 48 72 96 120 144

(e) 23° C (f) 26° C 20 20 15 15 10 10 5 5

48 96 144 192 24 48 72 96 120

(c)14° C (d) 19° C 20 20 15 15 10 10 Geometric mean maggot length (mm) 5 5

72 144 216 288 48 96 144 192 Fig. 2. The geometric mean length of lar- vae of the blowfly Calliphora vicina against ° ° (a)4 C (b) 10 C hours to a maximum of 5400 ADH. Each 20 20 data point represents the mean larval 15 15 length for a batch of (20) larvae reared in a single cup for a single temperature. (a) 10 10 30C, (b) 26C, (c) 23C, (d) 19C, (e) 14C, (f) 10C and (g) 4C. The fitted 5 5 hyperbolic function (solid line) with 95% confidence limits of the model (dashed 264 528 792 1056 144 288 432 576 lines) and for the observations (dotted Hours lines) are shown.

# 2006 The Authors Journal compilation # 2006 The Royal Entomological Society, Medical and Veterinary Entomology, 20, 106–114 Larval growth in Calliphora vicina 111

25

20

15

10

Fig. 3. The geometric mean length of lar- Geometric mean maggot length (mm) of top 75% ger larvae (upper three quartiles) of the 5 blowfly Calliphora vicina against accumu- lated degree hours (ADH) to a maximum of 5400 ADH. Each data point represents the mean larval length for the upper three quartiles of (20) larvae reared in a single cup. The fitted hyperbolic function (solid line) with 95% confidence limits of the 600 1200 1800 2400 3000 3600 4200 4800 5400 model (dashed lines) and for the observa- tions (dotted lines) are shown (r2 ¼ 0.943). Observed ADH discriminate between them. If the larvae are alive, those in This study emphasizes the importance of sampling entire the post-feeding phase can be identified through their beha- cohorts when measuring growth rates, rather than sequen- viour of actively moving away from the food source, no tially removing the largest individuals from just one batch longer feeding and by the way they contract into a barrel- of larvae. This is important because if the biggest – and shape when touched. If they are already dead, the crop of therefore the fastest growing – individuals within one batch the post-feeding larva (visible in dorsal view or by dissec- of larvae are removed at each sampling point then the study tion) is emptying or empty (Greenberg & Kunich, 2002). will be examining a progressively slower-growing popula- However, the confidence intervals of larvae of this age are tion of larvae. Furthermore, data obtained from just the extended (Fig. 3). upper range of larval sizes will have a smaller standard The feeding substrate can have a significant effect on deviation. This would lead to the reporting of spuriously larval growth rates. Preliminary studies showed the highest accurate results when calculating the larva’s age. The data growth rates on lamb liver, compared with beef mince, pork reported here indicate the full range of larval sizes (Fig. 1) mince and canned dog food. In other studies, liver has been that might be expected from individuals of the same age, shown to result in slower larval growth compared to but also focus on the longest 75% (Figs 3 and 4). growth on other organs (lung, kidney, heart and brain) This study has used field data to validate a model derived (Kaneshrajah & Turner, 2004; Clarke et al., 2005), but in from laboratory data. This is especially important given the latter, a different fly species (Lucilia sericata) was used. that there has been some concern that data derived from Maximum growth rates were also to be expected through larvae raised from single temperature rearings may differ in the choice of stocking level, selected as being the highest some minor, but fundamental way from those raised under density considered likely to avoid the possibilities of either fluctuating temperatures. Cyclic temperatures are more a rise in temperature through the formation of a larval mass natural but their effects on dipteran larval growth are (Turner & Howard, 1992) or food competition between ambiguous. In a number of instances, some species have larvae (Kamal, 1958; Williams & Richardson, 1983; been shown to develop more rapidly with fluctuating tem- Blackith & Blackith, 1984; Saunders et al., 1999). peratures than at a constant temperature, as long as the

# 2006 The Authors Journal compilation # 2006 The Royal Entomological Society, Medical and Veterinary Entomology, 20, 106–114 112 S. E. Donovan et al.

25

20

15

10

Fig. 4. The geometric mean length of lar- Geometric mean maggot length (mm) of top 75% 5 ger larvae (upper three quartiles) of the blowfly Calliphora vicina against accumu- lated degree hours (ADH) to a maximum of 2400 ADH. Each data point represents the mean larval length for the upper three quartiles of (20) larvae reared in a single cup. The fitted hyperbolic function (solid 600 1200 1800 2400 line) with 95% confidence limits of the model (dashed lines) and for the observa- Observed ADH tions (dotted lines) are shown (r2 ¼ 0.937). temperatures do not exceed the optimal range for the Evidence of biogeographical variation organism. In the case of Calliphora vomitoria (Linnaeus), this effect increased the larval growth rate by 7% The data presented here support the hypothesis that (Hagstrum & Hagstrum, 1970). Davies & Ratcliffe (1994) there is inherent biogeographical variation between popu- also found this to be the case, not only in C. vomitoria, but lations of C. vicina. Specifically, the following differences also in Protophormia terraenovae (Robineau-Desvoidy) and were apparent in comparison with Reiter’s (1984) work, Lucilia sericata (Meigen). By contrast, for C. vicina they which used Austrian flies. First, the present study found found that fluctuating temperatures tended to slow down complete and normal pupariation at temperatures below developmental times. Similarly, Greenberg (1991) found 16C; Reiter found that, below this temperature, the larvae that fluctuating temperatures tended to retard larval fell into a stationary phase, and only pupariated after the growth in four species – Chrysomya macellaria, temperature was again raised above this level. Second, a C. rufifacies, Phormia regina and (Lucilia) Phaenicia seri- lower lethal (high) temperature was determined; in this cata – although this effect was significant in only one of the study the larvae all died at 35C, whereas Reiter’s survived species, L. sericata. at this temperature, though they failed to pupariate nor- The timing of the temperature cycles can also be impor- mally. Many Calliphora spp. appear to be more cold adapted tant, because growth rates can change as the larvae pro- than other calliphorids (Faucherre et al., 1999). C. vicina gress through the various developmental stages. In the first has a distribution limited to areas where the summer and, to a lesser extent, the second larval instars the growth temperatures do not exceed 30C for sustained periods, but rate is slower (Williams, 1984). However, no significant the Austrian population may be adapted to the hotter, con- effect of fluctuating temperatures in the growth phase was tinental summer. Third, these data showed no temperature- suggested here, because the fit between the laboratory data related pattern of lower temperatures resulting in a larger and field data was good (Fig. 5). Consequently, it is reason- maximum size for the larvae. Finally, the results presented able to assume that fluctuating temperatures do not have a here indicated a lower minimum developmental tempera- significant effect on the overall ADH during the feeding ture for the U.K. population. A theoretical developmental phase for larvae of C. vicina. minimum was determined from Reiter’s data by calculating

# 2006 The Authors Journal compilation # 2006 The Royal Entomological Society, Medical and Veterinary Entomology, 20, 106–114 Larval growth in Calliphora vicina 113

2400

2160

1920

1680

1440

1200 Predicted ADH 960

Fig. 5. Regression of predicted accumu- 720 lated degree hours (ADH) against actual ADH for the field data of the blowfly Calliphora vicina larval growth, based on 480 the feeding phase only (maximum ADH of 2400), from a 1 : 1 fit. Four samples where the mean size exceeds the maximum 240 value provided by the size : ADH curve have been placed at the ADH where the maximum size is predicted. Filled circles 0 are estimated points; open circles are the 240 480 720 960 1200 1440 1680 1920 2160 points placed at the ‘maximum ADH’. Solid line is a 45 degree line. Observed ADH the best fit for the ADH values – calculated using the full collection of local data when applying growth curves to range of temperature/larval length data – again using a estimate larval age in specific forensic cases. range of minimum developmental temperatures, from 1C to 6C. The data fitted best using a developmental minimum for Austrian flies of between 4 C and 5 C Acknowledgements (R2 ¼ 0.889 and 0.892, respectively). Differences in experi- mental conditions and design between these studies mean We would like to thank Zoe Adams for valuable assistance that any comparisons of the results should be tentative. with laboratory work and maintaining the cultures. We However, the good fit between the laboratory and field also thank two anonymous reviewers for helpful comments data indicate robustness such that these comparisons have on the manuscript. This work was funded through a NERC some validity. Ideally, future studies should compare popu- small research grant (NER/A/S/ 2000/00311). lations from several different regions within the same experimental set-up. There is other evidence in the literature that would sup- References port the biogeographical variation hypothesis. There are significant differences between diapause initiation in popu- Adams, Z.J.O. & Hall, M.J.R. (2003) Methods used for the killing lations of C. vicina from southern England and from and preservation of blowfly larvae, and their effect on post- Finland (McWatters & Saunders, 1996, 1998). Saunders & mortem larval length. Forensic Science International, 138, 50–61. Hayward (1998) noted that there is geographical variation Ames, C. & Turner, B. D. (2003) Low temperature episodes in the in cold tolerance in C. vicina. Similarly, in L. sericata (syn. development of blowflies: implications for postmortem interval 17 Phaenicia sericata) Greenberg (1991) noted variation in estimation. Medical and Veterinary Entomology, , 178–186. Ash, N. & Greenberg, B. (1975) Developmental temperature responses growth in larvae from different continents: ADH of 8395 of the sibling species Phanicia sericata and Phanaecia pallescens. in Leningrad, Russia and 7684 in the mid West, U.S.A. Annals of the Entomological Society of America, 68, 197–200. (values representing development from larval hatch to the Blackith, R.E. & Blackith, R.M. (1984) Larval aggression in Irish point of pupariation). For C. vicina the range of ADH is flesh-flies (Diptera: Sarcophagidae). Irish Naturalists Journal, even higher (see above). This study emphasizes the need for 21, 237–257.

# 2006 The Authors Journal compilation # 2006 The Royal Entomological Society, Medical and Veterinary Entomology, 20, 106–114 114 S. E. Donovan et al.

Byrd, J.H. & Allen, J.C. (2001) Computer modeling of insect growth in Legal Investigations (ed. by J. H. Byrd and J. L. and its application to forensic entomology. Forensic Entomology: Castner), pp. 287–302. CRC Press, . the Utility of Arthropods in Legal Investigations (ed. By J. H. Byrd Kamal, A.S. (1958) Comparative study of thirteen species of sar- and J. L. Castner), pp. 303–330. CRC Press, Florida. cosaprophagous Calliphoridae and Sarcophagidae (Diptera) I. Byrd, J.H. & Butler, J.F. (1996) Effects of temperature on Bionomics. Annals of the Entomological Society of America, 51, Cochliomyia macellaria (Diptera: Calliphoridae) development. 261–271. Journal of Medical Entomology, 33, 901–905. Kaneshrajah, G. & Turner, B.D. (2004) Calliphora vicina larvae Byrd, J.H. & Butler, J.F. (1997) Effects of temperature on grow at different rates on different body tissues. International Chrysomya rufifacies (Diptera: Calliphoridae) development. Journal of Legal Medicine, 118, 242–244. Journal of Medical Entomology, 34, 353–358. McWatters, H.G. & Saunders, D.S. (1996) The influence of each Byrd, J.H. & Butler, J.F. (1998) Effects of temperature on parent and geographic origin on larval diapause in the blow fly, Sarcophaga haemorrhoidalis (Diptera: Sarcophagidae) develop- Calliphora vicina. Journal of Insect Physiology, 42, 721–726. ment. Journal of Medical Entomology, 35, 694–698. McWatters, H.G. & Saunders, D.S. (1998) Maternal temperature Byrd, J.H. & Castner, J.L. (2001) Insects of forensic importance. has different effects on the photperiodic response and the dura- Forensic Entomology: the Utility of Arthropods in Legal tion of larval diapause in blow fly (Calliphora vicina) strains Investigations (ed. By J. H. Byrd and J. L. Castner), pp. 43–79. collected at two latitudes. Physiological Entomology, 23,369–375. CRC Press, Florida. Me´ gnin, P. (1894) La faune des cadavres. Encyclope´die Clarke, K., Evans, L. & Wall, R. (2005) Growth rates of the Scientifiques Des Aides-Memoire. Gauthier-Villars et Fils, Paris. blowfly, Lucilia sericata, on different body tissues. Forensic Reiter, C. (1984) Zum wachstumsverhalten der maden der blauen Science International, (in press). schmeißfliege Calliphora vicina. Zeitschrift fu¨r Rechtsmedizin, 91, Davies, L. & Ratcliffe, G.G. (1994) Development rates of some 295–308. pre-adult stages in blowflies with reference to low temperatures. Saunders, D.S. & Hayward, S.A.L. (1998) Geographical and dia- Medical and Veterinary Entomology, 8, 245–254. pause-related cold tolerance in the blow fly, Calliphora vicina. Faucherre, J., Cherix, D. & Wyss, C. (1999) Behaviour of Journal of Insect Physiology, 44, 541–551. Calliphora vicina (Diptera: Calliphoridae) under extreme condi- Saunders, D.S., Wheeler, I. & Kerr, A. (1999) Survival and repro- tions. Journal of Insect Behaviour, 12, 687–690. duction of small blow flies (Calliphora vicina; Diptera: Grassberger, M. & Reiter, C. (2001) Effect of temperature on Calliphoridae) produced in severely overcrowded short day Lucilia sericata (Diptera: Calliphoridae) development with spe- larval cultures. European Journal of Entomology, 96, 19–22. cial reference to the isomegalen- and isomorphen-diagram. Sharpe, P.J.H. & DeMichele, D.W. (1977) Reaction kinetics of Forensic Science International, 120, 32–36. poikilotherm development. Journal of Theoretical Biology, 64, Grassberger, M. & Reiter, C. (2002a) Effect of temperature on 649–670. development of Liopygia (¼ Sarcophaga) argyrostoma Smith, K.E. & Wall, R. (1997) The use of as breeding sites (Robineau-Desvoidy) (Diptera: Sarcophagidae) and its forensic by the blowfly Lucilia sericata and other Calliphoridae. Medical implications. Journal of Forensic Science, 28, 565–577. and Veterinary Entomology, 11, 38–44. Grassberger, M. & Reiter, C. (2002b) Effect of temperature on Turner, B.D. & Howard, T. (1992) Metabolic heat generation in development of the forensically important holartic blow fly dipteran larval aggregations: a consideration for forensic ento- Protophormia terraenovae (R.-D.) (Diptera Calliphoridae). mology. Medical and Veterinary Entomology, 6, 179–181. Forensic Science International, 128, 177–182. Vinogradova, E.B. & Marchenko, M.I. (1984) The use of tempera- Grassberger, M., Friedrich, E. & Reiter, C. (2003) The blowfly ture parameters of fly growth in the medicolegal practice. Chrysomya albiceps (Wiedemann) (Diptera: Calliphoridae) as a Sudebno-Meditsinkskaya Ekspertiza, 27, 16–19. new forensic indicator in Central Europe. International Journal Wall, R., French, N.P. & Morgan, K.L. (1992) Effects of tempera- of Legal Medicine, 117, 75–81. ture on the development and abundance of the sheep blowfly Greenberg, B. (1991) Flies as forensic indicators. Journal of Lucilia sericata (Diptera: Calliphoridae). Bulletin of Medical Entomology, 28, 565–577. Entomological Research, 82, 125–131. Greenberg, B. & Kunich, J.C. (2002) Entomology and the Law – Wells, J.D. & Kurahashi, H. (1994) Chrysomya megacephala Flies as Forensic Indicators. Cambridge University Press, (Fabricius) (Diptera: Calliphoridae) development: rate, varia- Cambridge. tion and the implications for forensic entomology. Japanese Greenberg, B. & Tantawi, T.I. (1993) Different developmental Journal of Sanitary Zoology, 45, 303–309. strategies in two boreal blow flies (Diptera: Calliphoridae). Wells, J.D. & LaMotte, L.R. (1995) Estimating maggot age from Journal of Medical Entomology, 30, 481–484. weight using inverse prediction. Journal of Forensic Sciences, 40, Hagstrum, D.W. & Hagstrum, W.R. (1970) A simple device for 585–590. producing fluctuating temperatures, with an evaluation of the Williams, H. (1984) A model for the aging of fly larvae in forensic ecological significance of fluctuating temperatures. Annals of the entomology. Forensic Science International, 25, 191–199. Entomological Society of America, 63, 1385–1389. Williams, H. & Richardson, A.M.M. (1983) Life history responses Haskell, N.H., Lord, W.D. & Byrd, J.F. (2000) Collection of ento- to larval food shortages in four species of necrophagous flies mological evidence during death investigations. Forensic (Diptera: Calliphoridae). Australian Journal of Ecology, 8, Entomology: the Utility of Arthropods in Legal Investigations (ed. 257–263. by J. H. Byrd and J. L. Castner), pp. 81–120. CRC Press, Florida. Higley, L.G. & Haskell, N.H. (2000) Insect development and forensic entomology. Forensic Entomology: the Utility of Accepted 2 November 2004

# 2006 The Authors Journal compilation # 2006 The Royal Entomological Society, Medical and Veterinary Entomology, 20, 106–114