Factors affecting quality of temperature models for the pre-appearance interval of forensically useful insects

doi:10.1016/j.forsciint.2014.11.026

Forensic Science International

Volume 247, February 2015, Pages 28-35

https://doi.org/10.1016/j.forsciint.2014.11.026 Get rights and content

Highlights

•
Sampling frequency is the most important factor affecting quality of temperature models for PAI.
•
On-site, hourly temperature measurements give models of the best quality.
•
Number of carcasses in PAI studies may be largely reduced without compromising the quality of PAI models.
•
Small samples give good models, as long as the whole range of relevant temperatures is studied.

Abstract

In the case of many forensically important insects an interval preceding appearance of an insect stage on a corpse (called the pre-appearance interval or PAI) is strongly temperature-dependent. Accordingly, it was proposed to estimate PAI from temperature by using temperature models for PAI of particular insect species and temperature data specific for a given case. The quality of temperature models for PAI depends on the protocols for PAI field studies. In this article we analyze effects of sampling frequency and techniques, temperature data, as well as the size of a sample on the quality of PAI models. Models were created by using data from a largely replicated PAI field study, and their performance in estimation was tested with external body of PAI data. It was found that low frequency of insect sampling distinctly deteriorated temperature models for PAI. The effect of sampling techniques was clearly smaller. Temperature data from local weather station gave models of poor quality, however their retrospective correction clearly improved the models. Most importantly, current results demonstrate that sample size in PAI field studies may be substantially reduced, with no model deterioration. Samples consisting of 11–14 carcasses gave models of high quality, as long as the whole range of relevant temperatures was studied. Moreover, it was found that carcasses exposed in forests and carcasses exposed in early spring are particularly important, as they ensure that PAI data is collected at low temperatures. A preliminary best practice model for PAI field studies is given.

Introduction

Postmortem interval (PMI) or more frequently minimum PMI may be estimated from development or succession of insects on cadavers [1], [2], [3]. While estimating PMI from immature insects, it is convenient to divide it into the development interval and the pre-appearance interval (PAI) [4], [5]. Most recent progress in the field refers to methods for the estimation of the development interval [6], [7], [8], [9], [10], [11], [12], [13], [14], [15]. As for the PAI, substantial progress was achieved in the area of volatile organic compounds attracting insects to carcasses [16], [17], [18], [19], [20]. Moreover, recent works demonstrated that PAI of several forensically useful insects is dependent on temperature prevailing throughout this interval [21], [22], [23], [24]. PAI was regularly found to decrease exponentially with an increase in temperature, which indicates that an exponential decrease is a general model for the association of PAI with temperature in forensically significant insects [22], [24].

A recent progress in understanding the relation between PAI and temperature resulted in the proposition to estimate PAI from temperature [4], [5], [22], [23], [24]. These estimates require regression models capturing the dependence of PAI on average preceding temperatures in particular insect species. Such models were recently elaborated for many European species of Coleoptera [22], few European species of Diptera [24] and some Australian species of Diptera or Coleoptera [23]. The temperature models for PAI may be used in casework to estimate PAI simply by using temperatures specific for a given case. Such estimates of PAI may eventually be used to formulate estimates of PMI with a succession-based or a development-based approach [4], [5].

The temperature models for PAI may be created only from results of highly replicated pig carcass studies with a PAI-oriented design. Unfortunately, such studies are costly and time-consuming, so it would be useful to know how to effectively reduce their costs. Moreover, previous results suggest that differences in methods between various PAI studies may result in differences between resultant temperature models for PAI [4], [5], [22], [23]. Accordingly, a more profound understanding of factors affecting quality of temperature models for PAI is needed. Such knowledge will help to specify best practice in the PAI field studies, similarly to other fields of forensic entomology [25], [26], [27], [28], [29], [30].

It seems that the frequency with which insects are sampled during field studies is one of the most important factors. Sampling frequency was found to have a direct influence on the accuracy of temperature models for the development of forensically useful flies [26], [31]. It may have similar effects on temperature models for PAI. We predict that the accuracy of PAI data decreases with decreasing frequency of insect sampling, which, in turn, affects the quality of temperature models for PAI. However, the extent of these effects and in particular robustness of the models to decreasing frequency needs to be investigated.

Another factor of probable high relevance is techniques of insect sampling. Several techniques were used in forensic field carrion studies, i.e. manual collections (by hand or entomological net) [32], [33], [34], [35], [36], [37], [38], [39], pitfall traps [32], [35], [36], [37], [40], sticky traps [32], [38], Malaise traps [41] or Schoenly traps [42], [43], [44]. However, very few studies compared their performance in a forensic context. Schoenly et al. [32] demonstrated that the combination of pitfall traps and hand collections has the highest efficiency in catching forensically important taxa during field carcass experiments. It was also found that Schoenly traps are more efficient in catching adult flies than traditional manual collections [42]. It is however unclear whether and to what extent sampling techniques may influence the accuracy of PAI data and resultant temperature models for PAI. We predict that these effects exist and may have adverse influence on the quality of temperature models for PAI.

The quality of temperature data used in the model for PAI is also of probable high importance. On-site hourly measurements are preferable [22], [23], [24], however weather station data were also used and after retrospective correction gave surprisingly good models [4], [5]. Consequently, it would be interesting to determine how seriously weather station data deteriorate models, and to what extent retrospective correction of temperature improves the models. It is of relevance that corrections of weather station temperatures were found to improve the accuracy with which these temperatures represent temperatures of the place where corpse has been found [45]. However some results suggest that these corrections have uncertain benefit when estimating PMI from the development of flies [46].

The last relevant factor of the current study is the number of carcasses comprising the sample used for modeling purposes. Because temperature cannot be manipulated in PAI field studies, a good design for such studies is of key importance. A reliable model for PAI should cover a broad range of temperatures, and to attain this goal researchers should address some issues concerning experimental design. In order to study PAI at different temperatures, recent experiments involved several placements of carcasses, separated in time [22], [23], [24]. For the same purpose carcasses were exposed in different habitats [22], [24]. Moreover, carcasses may be distributed over different seasons, months and years. From this point of view several interesting questions arise. Firstly, is it more efficient to separate carcasses in time or space? Secondly, how should we separate carcasses in time or space to get the best range of temperatures? One may predict that both types of separation are important, however their effectiveness in collecting PAI data at different temperatures is probably different. Thirdly, what is the minimal number of carcasses that will give acceptable models? A recent work of Archer [23] revealed that for some insects quite good models may be created from as few as 10 carcasses, although their quality was worse than models from highly replicated studies [4], [5], [22], [24]. So we predict that number of carcasses in PAI fields studies may be reduced to some extent without sacrificing the quality of resultant models.

Section snippets

Data used in the analyses

All models were made by using PAI data from a large scale, PAI oriented pig carcass study. Details on the experimental design and protocols for handling of carcasses, sampling of insects and temperature measurements were specified by Matuszewski and Szafałowicz [22] and Matuszewski et al. [24]. Below there is just a brief description of the most important points. Thirty pig carcasses were separated in time: 18 April (four pigs), 15 June (six pigs), 4 July (six pigs), 21 July (four pigs), 16

Sampling frequency

There were significant differences between models in the fit, the c parameter and the accuracy of estimates (Table 2 and Fig. 1). Models at lower frequencies displayed poorer fit, a systematic increase in the c parameter and a substantial increase in the error of estimates (Fig. 1). Models at very low frequency lacked the inherent variability in PAI (Fig. 2 and for further comparison see models published by Matuszewski and Szafałowicz [22]).

Sampling techniques

Models significantly differed according to the fit and

Sampling frequency and techniques

As expected, reduction of sampling frequency resulted in evident deterioration of models; however the size of this effect was a surprise. These results suggest that sampling frequency is one of the most important factors affecting quality of temperature models for PAI in carrion insects. Richards and Villet [31] demonstrated that low frequency of sampling reduced the accuracy of K and D₀ calculations in a temperature model for the development of forensically important blowflies. Reduction of

Acknowledgements

Thanks are extended to K. Frątczak (Poznań, Poland), M. Jarmusz (Poznań, Poland), S. Konwerski (Poznań, Poland) and M. Szafałowicz (Warsaw, Poland) for help in field or laboratory work during 2011 or 2012 field studies, to K. Szpila (Toruń, Poland) for identification of adult and larval Stearibia nigriceps and to D. Bajerlein (Poznań, Poland) for identification of adult Saprinus semistriatus from 2012 experiment. We are also grateful to anonymous reviewers for their comments and suggestions,

References (47)

S. Matuszewski
Estimating the pre-appearance interval from temperature in Necrodes littoralis L. (Coleoptera: Silphidae)
Forensic Sci. Int.
(2011)
S. Niederegger et al.
Preliminary studies of the influence of fluctuating temperatures on the development of various forensically relevant flies
Forensic Sci. Int.
(2010)
D. Charabidze et al.
Larval-mass effect: characterisation of heat emission by necrophageous blowflies (Diptera: Calliphoridae) larval aggregates
Forensic Sci. Int.
(2011)
M. Mai et al.
Effect of different post-feeding intervals on the total time of development of the blowfly Lucilia sericata (Diptera: Calliphoridae)
Forensic Sci. Int.
(2012)
C.S. Richards et al.
Virtual forensic entomology: improving estimates of minimum post-mortem interval with 3D micro-computed tomography
Forensic Sci. Int.
(2012)
S.-W. Lin et al.
Life history data on the fly parasitoids Aleochara nigra Kraatz and A. asiatica Kraatz (Coleoptera: Staphylinidae), and their potential application in forensic entomology
Forensic Sci. Int.
(2013)
J.P. Michaud et al.
Predicting the visitation of carcasses by carrion-related insects under different rates of degree-day accumulation
Forensic Sci. Int.
(2009)
S. Matuszewski et al.
Temperature-dependent appearance of forensically useful beetles on carcasses
Forensic Sci. Int.
(2013)
S.C. Voss et al.
Annual and seasonal patterns of insect succession on decomposing remains at two locations in Western Australia
Forensic Sci. Int.
(2009)
S. Matuszewski et al.
Insect succession and carrion decomposition in selected forests of Central Europe. Part 3: succession of carrion fauna
Forensic Sci. Int.
(2011)

S. Matuszewski et al.

Insects colonising carcasses in open and forest habitats of Central Europe: search for indicators of corpse relocation

Forensic Sci. Int.

(2013)

S. Reibe et al.

How promptly do blowflies colonise fresh carcasses? A study comparing indoor with outdoor locations

Forensic Sci. Int.

(2010)

C. Prado e Castro et al.

Coleoptera of forensic interest: a study of seasonal community composition and succession in Lisbon, Portugal

Forensic Sci. Int.

(2013)

M. Battan Horenstein et al.

Seasonal structure and dynamics of sarcosaprophagous fauna on pig carrion in a rural area of Cordoba (Argentina): their importance in forensic science

Forensic Sci. Int.

(2012)

A.P. Johnson et al.

Experimental and casework validation of ambient temperature corrections in forensic entomology

J. Forensic Sci.

(2012)

J.D. Wells et al.

Estimating the postmortem interval

J. Amendt et al.

Forensic entomology: applications and limitations

Forensic Sci. Med. Pathol.

(2011)

M.H. Villet et al.

Advances in entomological methods for death time estimation

S. Matuszewski

Estimating the pre-appearance interval from temperature in Creophilus maxillosus L. (Coleoptera: Staphylinidae)

J. Forensic Sci.

(2012)

M.B. Gallagher et al.

Variation in developmental time for geographically distinct populations of the common green bottle fly, Lucilia sericata (Meigen)

J. Forensic Sci.

(2010)

H. Fremdt et al.

Diapause-specific gene expression in Calliphora vicina (Diptera: Calliphoridae) – a useful diagnostic tool for forensic entomology

Int. J. Leg. Med.

(2014)

M. Baque et al.

Strengthen forensic entomology in court-the need for data exploration and the validation of a generalised additive mixed model

Int. J. Leg. Med.

(2013)

P. Boehme et al.

The analysis of temporal gene expression to estimate the age of forensically important blow fly pupae: results from three blind studies

Int. J. Leg. Med.

(2014)

Cited by (15)

Estimating crime scene temperatures from nearby meteorological station data
2020, Forensic Science International
Citation Excerpt :
Insects are poikilothermic and, therefore, the role of temperature in forensic entomology applications, particularly its effects on rates of development, have been well emphasised [e.g. 1,7–9]. Temperature is a fundamental influencing factor in body decomposition as well as for estimations of the pre-appearance interval (PAI) in insect succession patterns [7,10,11]. Seasonal temperature variations can influence body decomposition rates, as reported by a study in 2004, which found that decomposition rates increased in higher temperatures and with greater rainfall [9].
The importance of temperature data in minimum postmortem interval (_minPMI) estimations in criminal investigations is well known. To maximise the accuracy of _minPMI estimations, it is imperative to investigate the different components involved in temperature modelling, such as the duration of temperature data logger placement at the crime scene and choice of nearest weather station to compare the crime scene data to. Currently, there is no standardised practice on how long to leave the temperature data logger at the crime scene and the effects of varying logger duration are little known. The choice of the nearest weather station is usually made based on availability and accessibility of data from weather stations in the crime scene vicinity. However, there are no guidelines on what to look for to maximise the comparability of weather station and crime scene temperatures.
Linear regression analysis of scene data with data from weather stations with varying time intervals, distances, altitudes and microclimates showed the greatest goodness of fit (R²), i.e. the highest compatibility between datasets, after 4–10 days. However, there was no significant improvement in estimation of crime scene temperatures beyond a 5-day regression period. The smaller the distance between scene and weather station and the higher the similarity in environment, such as altitude and geographical area, resulted in greater compatibility between datasets.
Overall, the study demonstrated the complexity of choosing the most comparable weather station to the crime scene, especially because of a high variation in seasonal temperature and numerous influencing factors such as geographical location, urban ‘heat island effect’ and microclimates. Despite subtle differences, for both urban and rural areas an optimal data fit was generally reached after about five consecutive days within a radius of up to 30 km of the ‘crime scene’. With increasing distance and differing altitudes, a lower overall data fit was observed, and a diminishing increase in R² values was reached after 4–10 consecutive days. These results demonstrate the need for caution regarding distances and climate differences when using weather station data for retrospective regression analyses for estimating temperatures at crime scenes. However, the estimates of scene temperatures from regression analysis were better than simply using the temperatures from the nearest weather station. This study provides recommendations for data logging duration of operation, and a baseline for further research into producing standard guidelines for increasing the accuracy of _minPMI estimations and, ultimately, greater robustness of forensic entomology evidence in court.
Post-mortem interval estimation based on insect evidence in a quasi-indoor habitat
2019, Science and Justice
Citation Excerpt :
Accordingly, we used the regression equation to retrospectively correct temperatures from the station to be used in PMI estimation. In general, insect age was estimated using thermal summation method [20–22], and supplemented with the PAI estimated using temperature methods [23, 24]. Because larval PAI may be estimated with the average error of about 20% [23] and because there is no contemporary data on the error rate of insect age estimates (it is probably lower than 20%), we decided to include a conservative 20% error rate for both age and PAI estimates.
Insects collected on indoor cadavers are frequently used for post-mortem interval (PMI) estimation. Buildings encountered during crime investigations vary according to temperatures inside, the extent of insect access restriction or sanitary conditions. This article reports the PMI oriented analyses of insect evidence sampled from the human cadaver in the atypical indoor habitat. The body was found in the uninhabited house, on the floor covered with rubbish, in the room with no doors and windows. Thermal conditions in the room were less variable than in the local weather station, however still much more variable compared to the typical indoor habitat, indicating the need for retrospective correction of temperature records from the station. Cadaver entomofauna was surprisingly diverse and abundant. We recorded several taxa usually not occurring on indoor cadavers, e.g. immature stages of Necrodes littoralis (Coleoptera: Silphidae) or Stearibia nigriceps (Diptera: Piophilidae). PMI was based on the age and the pre-appearance interval estimated for live puparium of S. nigriceps, giving the total interval of 37 (±7.4) days plus 4–20 days resulting from the absence of first colonizing specimens of the species. This estimate was corroborated with the age estimate for empty puparia of Sarcophaga argyrostoma (Diptera: Sarcophagidae) with traces of Nasonia sp. (Hymenoptera: Pteromalidae) eclosion. Other insects indicated shorter but consistent PMI. Difficulties and limitations of insect-based PMI estimations in unusual indoor habitats are discussed.
Estimation of postmortem interval (PMI) based on empty puparia of Phormia regina (Meigen) (Diptera: Calliphoridae) and third larval stage of Necrodes littoralis (L.) (Coleoptera: Silphidae) – Advantages of using different PMI indicators
2018, Journal of Forensic and Legal Medicine
Citation Excerpt :
Various tools were recently developed in forensic entomology to estimate the age of insects sampled on a cadaver.3–8 Moreover, methods of PAI estimation from temperature data were developed for different species of forensically-useful insects.9,10 Much attention is also paid to studies on insect succession on carrion in relation to various cadaver treatments, body mass, geographical region, season and habitat of decomposition.11–17
On 16 July 2015, a body of a 64-year-old man in advanced decomposition was found in an open area of the suburb of Śrem (western Poland). Postmortem interval (PMI) was estimated by forensic pathologist for 3–6 weeks. Insects were sampled from the cadaver and the soil from below the cadaver. Empty puparia of Phormia regina were the most developmentally advanced specimens of blowflies. Moreover, third instar larva of Necrodes littoralis was collected directly from the cadaver. For the estimation of minimum PMI from puparia of P. regina, thermal summation method was used to estimate the total immature development interval of this species. In the case of larval N. littoralis, the pre-appearance interval (PAI) was estimated using temperature method and the development interval (DI) using thermal summation method. Average daily temperatures from the nearby weather station were used, as well as the weather station temperatures corrected by 1 °C and 2 °C. The estimates were as follows: 36–38 days using empty puparia of P. regina and 37–40 days using larva of N. littoralis (for the uncorrected temperatures), 31–34 days using both P. regina and N. littoralis (temperatures corrected by +1 °C), 24–27 days using P. regina and 28–29 days using N. littoralis (temperatures corrected by +2 °C). It was concluded that death occurred 24–40 days before the body was found and most probably 24–34 days before the body was found. This is the first report when PMI was approximated by the age estimates combined with the PAI estimates. Moreover, the case demonstrates the advantages of using different entomological indicators and an urgent need for the more robust developmental model for N. littoralis, as it proved to be highly useful for the estimation of PMI.
The importance of Saprinus semistriatus (Coleoptera: Histeridae) for estimating the minimum post-mortem interval
2018, Legal Medicine
Citation Excerpt :
This is not in accordance with our study where the PAI increased with increasing temperature, ranging from 3 to 14 days at temperatures from 13.7 to 16.6 °C (Fig. 3), although the number of data points measured within this small temperature range between the two studies is almost the same (this study n = 10, n = 12 [5]). The discrepancy between the two studies may be due to the fact that our study did not meet all the criteria suggested by Matuszewski and Madra [24] to assure the quality of the temperature model for PAI. Although the frequency of insect sampling and the combination of pitfall traps and hand collections was carried out, one of the most important factors, i.e. the broad range of temperature data that should be covered [24], was not realized in our study.
In forensic science, the use of entomological evidence to estimate the minimum post-mortem interval can be crucial. However, not all cadaver-visiting insects are equally useful. Our focus is on the histerid beetle Saprinus semistriatus (Scriba 1790) (Histeridae; Coleoptera). Histeridae are common predators that feed mainly on dipteran larvae on carrion and dung. We review 23 publications mentioning this species and provide new experimental data on its temporal pattern beneath and on hanging pig cadavers. In a field experiment near Neuchâtel, Switzerland, we recorded the abundance of S. semistriatus on ten decomposing pig cadavers (Sus scrofa) over a 32-day period in summer 2013 (July, 01- August; 02). Five cadavers were placed on the ground and five cadavers were hung one metre above the ground. Insects were collected from pitfall traps and by manual sampling. The abundance of S. semistriatus was significantly higher during the active stage than during the fresh and the bloated stages of decomposition in both, ground and hanging pigs. However, S. semistriatus was more abundant on the ground than on the hanging cadavers. The literature and our new data show that S. semistriatus is present on cadavers during a relatively short period of time (approximately two weeks), mainly during the active decay stage, but it may also occasionally occur in the bloated stage. Identifying key indicators such as S. semistriatus can help optimise forensic research by focusing on the most informative taxa. A few key indicators for each decomposition stage may constitute an optimal toolbox for forensic entomologists.
Optimising crime scene temperature collection for forensic entomology casework
2017, Forensic Science International
The value of minimum post-mortem interval (minPMI) estimations in suspicious death investigations from insect evidence using temperature modelling is indisputable. In order to investigate the reliability of the collected temperature data used for modelling minPMI, it is necessary to study the effects of data logger location on the accuracy and precision of measurements. Digital data logging devices are the most commonly used temperature measuring devices in forensic entomology, however, the relationship between ambient temperatures (measured by loggers) and body temperatures has been little studied. The placement of loggers in this study in three locations (two outdoors, one indoors) had measurable effects when compared with actual body temperature measurements (simulated with pig heads), some more significant than others depending on season, exposure to the environment and logger location. Overall, the study demonstrated the complexity of the question of optimal logger placement at a crime scene and the potential impact of inaccurate temperature data on minPMI estimations, showing the importance of further research in this area and development of a standard protocol. Initial recommendations are provided for data logger placement (within a Stevenson Screen where practical), situations to avoid (e.g. placement of logger in front of windows when measuring indoor temperatures), and a baseline for further research into producing standard guidelines for logger placement, to increase the accuracy of minPMI estimations and, thereby, the reliability of forensic entomology evidence in court.
Heat accumulation and development rate of massed maggots of the sheep blowfly, Lucilia cuprina (Diptera: Calliphoridae)
2016, Journal of Insect Physiology
Citation Excerpt :
Blowflies (Diptera: Calliphoridae) are useful indicators of the time that has elapsed since a carcass or corpse was exposed to colonization by insects (Higley and Haskell, 2010; Villet et al., 2010; Rivers and Dahlem, 2014). Gravid females may arrive at carrion within minutes of death, and will lay eggs promptly under optimal conditions (Matuszewski and Madra, 2015). Larvae can hatch from 6 h onwards, depending on ambient temperatures and precocial fertilisation (Erzinçlioğlu, 1990; Wells and King, 2001), and then feed on the carcass until they reach the post-feeding stage in preparation for pupation (Villet et al., 2010; Rivers and Dahlem, 2014).
Blowfly larvae aggregate on exposed carcasses and corpses and pass through three instars before wandering from the carcass and pupating. The developmental landmarks in this process can be used by forensic entomologists to estimate the time since the insects colonised the carcass, which sets a minimum post mortem interval. Large aggregations of feeding larvae generate a microclimate with temperatures up to 15 °C above ambient conditions, which may accelerate larval development and affect forensic estimates of post-mortem intervals. This study investigated the effects of heat accumulated by maggot masses of Lucilia cuprina at aggregations of 20, 50 and 100 larvae, each at incubation temperatures of 18 °C, 24 °C and 30 °C, using body length and life stage as developmental indicators. Aggregation temperatures reached up to 18.7 °C above ambient temperature, with significant effects of both size and temperature of the aggregation on the development time of its larvae. Survivorship was highest for all life stages at 24 °C, which is near the developmental optimum of L. cuprina. The results of this study provide a broadly applicable method of quantifying heat accumulation by aggregations of a wide range of species of forensic importance, and the results obtained from such studies will demonstrate that ambient temperature cannot be considered the only source of heat that blowfly larvae experience when they develop on a carcass. Neglect of temperatures within larval aggregations will result in an overestimation of post-mortem intervals and thus have far-reaching medicolegal consequences.

View all citing articles on Scopus

View full text