ARTHROPOD BIOLOGY

Selected Life-History Traits of Black Soldier Flies

(Diptera: Stratiomyidae) Reared on Three Artificial Diets
JEFFERY K. TOMBERLIN,1 D. CRAIG SHEPPARD, AND JOHN A. JOYCE
Department of Entomology, University of Georgia, Tifton, GA 31793

Ann. Entomol. Soc. Am. 95(3): 379Ð386 (2002)

ABSTRACT Hermetia illucens (L.) was reared on three larval diets to determine their effects on
preimaginal development and selected adult life-history traits. Prepupal and adult characteristics were
examined for individuals reared on each diet and compared with Þeld-collected prepupae and
corresponding emergent adults. Diet did not signiÞcantly inßuence development or survivorship to
the prepupal stage. However, adult emergence for all diets was signiÞcantly less than that determined
for the wild population. Development time from egg to adult for individuals reared on the diets at 27⬚C
ranged from 40 to 43 d with the larval stage lasting 22Ð24 d. We observed ⬎96% larval survivorship
to the prepupal stage and 21Ð27% adult emergence. Females accounted for 55Ð 60% of emergent adults
across treatments. Specimens reared on each diet were reduced in size, longevity, and calorie content
in comparison to specimens from the wild population. Males within diet treatments and Þeld-collected
specimens were signiÞcantly smaller than females and emerged 1Ð2 d before females. Additionally,
males reared on the diets and provided water lived for 9 d, whereas females lived for 8 d. This
information indicates the diets might be used for rearing soldier ßies. However, further reÞnement
is needed to produce adults similar to those found in nature.

KEY WORDS Diptera, Stratiomyidae, Hermetia illucens

THE BLACK SOLDIER FLY, Hermetia illucens (L.), is a large 1981) and swine (Newton et al. 1977). Prepupae,
(13 to 20 mm), wasp-like ßy (May 1961). It has three when dried, have an estimated value comparable to
generations per year in the southeastern United States menhaden Þsh meal. If used live, as specialty feed, or
and can be collected from late spring through early fall marketed to exploit its other unique qualities (i.e.,
(Sheppard et al. 1994). Larvae occur in assorted de- essential fatty acids and chitin), the value of the prod-
composing materials, such as fruits, animals, and ma- uct might be higher (Sheppard et al. 1994). Addition-
nure (James 1935). ally, a system has already been developed for self-
The black soldier ßy is interesting because its pres- harvesting the larvae by directing their search for
ence can be used to solve many of the problems as- pupation sites into collection bins (Sheppard et al.
sociated with large manure accumulations at conÞned 1994).
animal feeding operations. House ßy, Musca domestica This manure management system depends on a ro-
L., control due to manure being colonized by the black bust soldier ßy population for dependable inoculation
soldier ßy has been reported by Furman et al. (1959), of the manure with larvae. Presently, little is known
Tingle et al. (1975), Sheppard (1983), and Axtell and about the biology of H. illucens. Rearing of the black
Arends (1990). Additionally, Tingle et al. (1975) doc- soldier ßy has been attempted in the past (Tingle et al.
umented that black soldier ßy larvae reduce waste 1975) but was unsuccessful until 2002 (Sheppard et al.
within poultry facilities. Sheppard et al. (1994) fol- 2002). With the development of suitable methods, the
lowed up their work and determined that the black soldier ßy can now be produced and made available as
soldier ßy can reduce manure accumulation by 42Ð a biological control and waste management agent.
56%. Concentrations of nitrogen, and other nutrients The primary objective of this study was to deter-
are also signiÞcantly lower in this reduced manure mine basic biological parameters for the black soldier
(D.C.S. unpublished data), thus further reducing the ßy when reared in selected diets. Additionally, we
potential for pollution. compared life-history data for prepupae and adults
Soldier ßy prepupae can be used as feed for a variety reared in these diets to data recorded for Þeld-col-
of animals, including Þsh (Bondari and Sheppard lected prepupae and the emergent adults. Information
from this study is important for improving rearing
1
Current address: Department of Biological and Agricultural En-
methods necessary for integrating the black soldier ßy
gineering, University of Georgia, P.O. Box 748, Tifton, GA 31793 into current waste management strategies in livestock
(e-mail: [email protected]). and poultry facilities.

0013-8746/02/0379Ð0386$02.00/0 䉷 2002 Entomological Society of America

380 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 95, no. 3

Table 1. Composition (%) of diets used to rear black soldier ture levels of 70%, whereas the CSMA diet moisture
flies was slightly lower at 64%. For each diet during an
Gainesville Layer hen
experiment, 300 4-d-old soldier ßy larvae were placed
Constituent CSMA in each of three 1-liter clear plastic cups. Cups were
diet ration
a each covered with a paper towel and held in the
Alfalfa meal 30.0 27.0
Wheat bran 50.0 33.0 a rearing room. Feeding was terminated in a treatment
Corn meal 20.0 Ñ a
when a cumulative 40% of the larvae in the three cups
BrewersÕ dried grain Ñ 40.0 Ñ reached the prepupal stage. However, daily observa-
Factors
Protein 15.3 19.0 15.0
tions continued until all larvae had entered the pre-
Fat 3.8 3.0 3.0 pupal stage or died. Prepupae were identiÞed by a
Fiber 12.6 20.0 5.0 change in integument color from larval white to black
Ash 6.3 8.0 13.7 (May 1961).
Calcium 4.9 4.0 5.1
Immature Life-History Traits. Larval weight in
a
Proprietary information, but comparable to that placed in other each treatment was determined daily by selecting 10
treatments. larvae from each cup per treatment and weighing
them individually on a Mettler AT261 DeltaRange
balance (Mettler-Toledo, Hightstown, NJ). After
Materials and Methods weighing, larvae were returned to their respective
cups. Final larval weight was deÞned as the average
Acquisition of Flies. At a 100 bird caged-layer poul- larval weight recorded on the day when 40% of all
try house at the Coastal Plain Experiment Station in individuals in a treatment reached the prepupal stage.
Tifton, GA, we collected 10 Ð20 female soldier ßies Prepupae were removed daily and transferred to
probing the manure surface with their ovipositors and another 1-liter cup (containing 100 g white sand as a
placed them in a 0.5-liter glass jar with a 2 by 3-cm roll pupation medium) designated to hold prepupae from
of corrugated cardboard as an oviposition substrate that particular replicate of a diet treatment. Addition-
(Booth and Sheppard 1984). After 24 h, the cardboard ally, each day, three of the prepupae collected from
containing soldier ßy eggs was placed in another 0.5- each replicate of each treatment were weighed and
liter Sweetheart plastic cup (Sweetheart Cup Com-
placed in petri plates and stored at ⫺15⬚C for later
pany, Chicago, IL) covered with a dry paper towel,
determination of caloric content. Cups containing
and stored in a rearing room (27⬚C, 60 Ð70% RH, and
prepupae were covered with 50 by 50 cm of mosquito
a photoperiod of 14:10 [L:D] h) until larvae emerged
netting held in place with a rubber band, placed in the
4 d later. Once larvae were observed, 50 g of a 15%
rearing chamber, and monitored daily for adult emer-
protein laying hen ration (Country Acres Feed, Brent-
gence. Percentage of larvae reaching the adult stage
wood, MO) mixed with 90 ml of water (70% moisture)
was placed in the container. The larvae were allowed was recorded for each diet treatment. This could not
to feed on the medium for 4 d before being used in the be determined for the wild population, which was
experiment. sampled when ßies reached the prepupal stage. Days
Wild Fly Population. Black soldier ßy prepupae from egg collection to adult emergence were recorded
were collected from swine and poultry units located in for each ßy reared on the diets.
Tifton, GA, and held in the above mentioned rearing Prepupal Calorie Content. Mean caloric content of
room in six 1-liter clear Sweetheart plastic cups at a prepupae per treatment was determined using bomb
density of ⬇500 per cup. Prepupal and adult charac- calorimetry (combustion) techniques (Dutcher 1983).
teristics compared across diet treatments were com- Frozen prepupae were held from seven to 14 d in a
pared with data recorded for the wild population (see convection oven set at 55⬚C. Individual dried prepu-
below). pae were then weighed on a Mettler EA 100 electronic
Experimental Design. The following three diets balance (Metler-Toledo), ground to ⬇1-mm pieces,
were evaluated (Table 1): (1) Gainesville diet, which mixed with calorimeter grade benzoic acid (Sigma,
was developed for rearing house ßies (Hogsette 1985) St. Louis, MO) (1:10 specimen weight: benzoic acid
and the diet currently suggested for mass rearing the weight) and compressed into a pellet. We were unable
black soldier ßy (Sheppard et al. 2002); (2) CSMA, an to make pellets with prepupae collected from the wild.
artiÞcial rearing media for house ßy larvae (Chemical These prepupae, unlike those reared on the diet treat-
Specialties ManufacturersÕ Association, Ralston Pu- ments, appeared to contain greater amounts of fats and
rina, St. Louis, MO); (3) the 15% protein layer hen attempts to crush oven-dried specimens resulted in
ration, which had been used in preliminary trials for observable loss of materials. Therefore wild prepupae
rearing H. illucens. were analyzed without being mixed with benzoic acid.
The experiment was replicated three times from To ensure that this difference in techniques did not
April through October 1999. Preliminary studies using inßuence the results, caloric content was determined
the layer hen ration rate of 10 g mixed with 17 ml water for prepupae from each diet treatment prepared with
to 300 larvae daily resulted in the heaviest Þnal larval and without benzoic acid. A Parr oxygen bomb calo-
weight. Therefore, this rate was selected as the feeding rimeter (model No. 1341) (Parr Instrument, Moline,
rate to be used for all three diets. The layer hen ration IL) was used to determine caloric content of pellets.
and Gainesville diet mixed with water had initial mois- Caloric content per mg ash-free dry weight per pre-
May 2002 TOMBERLIN ET AL.: SELECTED LIFE-HISTORY TRAITS FOR BLACK SOLDIER FLIES 381

pupa was determined for each diet treatment and for 5-liter bucket (Plastic Packaging, West SpringÞeld,
the wild population. MA) containing ⬇1 kg Gainesville diet saturated with
Adult Life-History Traits. Weight, sex, and time water was placed on a cinder block (40 cm high) in
from oviposition to adult emergence were recorded each cage to attract ovipositing individuals. Methods
for each adult from the diet treatments. Individual for collecting eggs were adapted from Booth and
adults were placed in 35-ml Solo cups (Solo, Urbana, Sheppard (1984). Three layers of corrugated card-
IL) capped with a breathable lid. Half of the adult ßies board (each ßute opening measured 1 by 4 mm) were
were provided 0.125 ml water daily via a 1.25-cm glued together, cut into 5.0 by 2.5 by 1.0-cm blocks,
needle inserted through the lid. The remaining ßies and taped inside the 5-liter bucket ⬇3 cm above the
were held in similar conditions, but without water. media. These cardboard blocks were replaced daily.
Longevity was recorded for both groups. An egg clutch was deÞned as a ßute containing ⬎100
Egg Development and Clutch Size. Initially, we eggs. Egg clutches collected each day from each treat-
hypothesized that larval media inßuences the number ment were weighed and stored in vials containing 80%
of eggs produced by black soldier ßies and that mating ethanol until the eggs could be counted.
was not required to produce eggs. Therefore, to test Egg Collection System. We suspected that eggs per
this hypothesis and determine the number of eggs per ßute did not accurately represent the number of eggs
individual, a sample of nonmated females from each oviposited per individual. Therefore, we conducted an
diet treatment and from the wild population was dis- experiment to evaluate the relationship between ßute
sected in saline solution in a petri plate 1Ð 4 d after dimensions and eggs deposited per ßute by black sol-
emergence. Ovaries were removed and the number of dier ßies reared in our colony and held in a 2 by 2 by
developed eggs recorded. Specimens were dissected 3-m cage placed in the greenhouse. We examined
and examined under a Leica2000 Zoom dissecting mi- three cardboard ßute sizes; one with a smaller opening
croscope (Leica, Buffalo, NY). A fully developed egg (1 by 3 mm); one with the same opening (1 by 4 mm);
was deÞned as elongate and creamy to white in color, and the third with a larger opening (2 by 5 mm) than
while an oocyte was deÞned as spherical and trans- that used in the diet experiments. Cardboard blocks
parent. Oviposition by females in 35-ml cups was also were collected and replaced daily. Eggs from each
recorded to determine if the number of eggs in dis- ßute were stored in a vial containing 80% ethanol. Egg
sected specimens was similar to the number deposited number per ßute was determined as described above.
by females reared on the same diet. Statistical Analysis. Procedure GLM (SAS Institute
Dissected, unmated specimens did not contain de- 1992) was used to analyze the recorded data. Least
veloped eggs. Therefore, we suspected that mating signiÞcant difference (LSD) test (SAS Institute 1992)
was required for egg production. To test this hypoth- was used following a signiÞcant F test (P ⱕ 0.05) to
esis, ⬇500 (⬍24 h postemergence) black soldier ßies separate mean differences between diet treatments,
were collected from a colony (larvae fed Gainesville and the wild population for: percent survival to adult
diet) and released at 1500 hours in a 2 by 2 by 3-m cage stage, percent females and males to emerge, pupal skin
in a greenhouse maintained at ⬇30⬚C. Observations weights, adult male and female weights, and male and
and collection of mating pairs were made hourly be- female longevity with and without water. Additionally
ginning at 1600 hours on the day the adults were for the diet treatments only, the previously described
released and continued daily from 0800 to 1700 hours. procedure was used to separate mean larval and pupal
Fifty-Þve mating pairs were collected by the conclu- development periods and Þnal larval weight. Percent-
sion of the second day. Mated females were weighed age values were normalized by arcsine transformation.
within 1 h of mating, placed in capped 35-ml cups, and Mean egg number and total weight of eggs per ßute
held in the rearing room and provided water as pre- were analyzed using the procedures previously de-
viously described. We determined in a previous ex- scribed. Regression and PearsonÕs correlation analysis
periment that oviposition generally occurred 2 d after (SAS Institute 1992) were used to determine the re-
mating (Tomberlin 2001). Therefore, 2 d after mating, lationship between egg number and mass weight per
the ovaries of 45 mated females were dissected and ßute for each treatment, and adult weight and lon-
percent with developed eggs determined. Number of gevity.
developed eggs per individual was also recorded.
The remaining 10 females were allowed to oviposit
Results
in corrugated cardboard taped to the opening of the
35-ml cup. These eggs were removed and the number Immature Life-History Traits. Final larval weight
determined based on mass weight (Booth and Shep- across treatments ranged from 0.153 to 0.171 g (Table
pard 1984). The estimated number of eggs deposited 2) and did not differ signiÞcantly (F ⫽ 2.35; df ⫽ 2, 2;
in the cardboard was compared with the number of P ⬎ 0.2118). Survivorship of larvae to the prepupal
eggs recorded for dissected mated females. stage was not signiÞcantly different across diet treat-
To determine the number of eggs oviposited by ments (F ⫽ 1.07; df ⫽ 2, 23; P ⬎ 0.3592) and ranged
adults reared on each of the three diets and from the from 96.0 to 97.8% (Table 3). Time from egg to pre-
wild population, we released ⬇500, ⬍24 h old, adults pupal stage per treatment ranged from 22.5 to 24.1 d
from each treatment into 2 by 2 by 3-m cages placed (Table 4) and did not differ signiÞcantly (F ⫽ 0.74;
in the greenhouse previously described. Each treat- df ⫽ 4, 4; P ⬎ 0.6121). Mean prepupal weights differed
ment had a separate cage. To collect egg masses, a signiÞcantly among diet treatments including the wild
382 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 95, no. 3

Table 2. Comparison of larval, prepupal, pupal skin, and adult weights (grams), and prepupal caloric content per milligram of ash-free
dry weight for H. illucens reared on three diets and from field-collected prepupae

Prepupal Pupal skin Adult weight

Diet Final larval weight Caloric content/mg
weight weight Male Female
Gainesville diet 0.157 ⫾ 0.077A 0.104 ⫾ 0.027A 3.51 ⫾ 0.28A 0.011 ⫾ 0.003A 0.046 ⫾ 0.005A,a 0.056 ⫾ 0.005A,b
n⫽3 n ⫽ 30 n ⫽ 23 n ⫽ 40 n ⫽ 272 n ⫽ 339
CSMA 0.153 ⫾ 0.063A 0.107 ⫾ 0.041A 4.21 ⫾ 0.39AB 0.011 ⫾ 0.002A 0.044 ⫾ 0.005A,a 0.054 ⫾ 0.005A,b
n⫽3 n ⫽ 30 n ⫽ 22 n ⫽ 45 n ⫽ 254 n ⫽ 320
Layer hen ration 0.171 ⫾ 0.043A 0.111 ⫾ 0.034A 4.48 ⫾ 0.27B 0.016 ⫾ 0.006B 0.053 ⫾ 0.007B,a 0.064 ⫾ 0.006B,b
n⫽3 n ⫽ 30 n ⫽ 20 n ⫽ 45 n ⫽ 234 n ⫽ 316
Wild population NA 0.220 ⫾ 0.040B 5.95 ⫾ 0.32C 0.025 ⫾ 0.008C 0.085 ⫾ 0.016C,a 0.111 ⫾ 0.022C,b
n ⫽ 30 n ⫽ 15 n ⫽ 30 n ⫽ 57 n ⫽ 70

Means within a column followed by different capital letters are signiÞcantly different. Means for both sexes with different lowercase letters
within a treatment differ signiÞcantly (P ⱕ 0.05; LSD, SAS Institute 1992). NA, Not applicable.

population (F ⫽ 146.79; df ⫽ 3, 116; P ⬍ 0.0001) and soldier ßies reared on diets were signiÞcantly smaller
ranged from 0.104 to 0.220 g (Table 2). Pupal skins than females (F ⫽ 493.59; df ⫽ 7, 1854; P ⱕ 0.0001)
accounted for ⬇10% of their weight (Table 2). Pre- reared on the same diet. This size differential also
pupae reared on layer hen ration had the greatest occurred in the wild population (Table 2). However,
numerical weight among diet treatments, but still when comparing the same sex across treatments, wild
weighed signiÞcantly less than prepupae from the wild males and females were signiÞcantly larger than their
populations. Prepupal skin weight differed signiÞ- diet-reared counterparts. Although not signiÞcantly
cantly between diet treatments (F ⫽ 75.47; df ⫽ 3, 156; different (F ⫽ 9.92; df ⫽ 7, 10; P ⬎ 0.2875), adult males
P ⱕ 0.0001) with skins from specimens reared on layer generally emerged earlier than females (Table 4).
hen ration and from the wild population weighing the An interaction effect was determined for larval diet
most. and the provision of water on adult longevity (F ⫽
Prepupal Calorie Content. Mean calorie content 48.22; df ⫽ 15, 1,764; P ⱕ 0.0001) of black soldier ßies
per mg ash-free dry weight per prepupa was signiÞ- (Table 4). Adult males reared on a diet and provided
cantly different across treatments (F ⫽ 8.70; df ⫽ 3, 76; water as adults lived 9.3Ð9.7 d, whereas those not
P ⱕ 0.0001) and ranged from 3.51 to 5.95 cal per mg provided water lived 6.0 Ð7.1 d. Females reared on the
ash-free dry weight, with those specimens collected treatments and provided water as adults lived 7.9 Ð 8.5
from the wild having the greatest calorie content (Ta- d, whereas those not provided water lived 6.1Ð 6.4 d.
ble 2). Wild prepupae weighed more and had a higher Males and females that emerged from wild-collected
concentration of fat, yielding ⬇3.5 times more calo- prepupae lived signiÞcantly longer than those reared
ries, than prepupae reared on the diet treatments. on the artiÞcial diets. Wild males and females lived
Prepupae reared on the diets contained ⬇365 calories, ⬇14 d when provided water as adults and 7.8 and 8.2 d,
whereas the wild prepupae contained an average of respectively, when not provided water. Male and fe-
1,309 calories. No signiÞcant differences were deter- male longevity within a treatment when not provided
mined for the caloric content of prepupae prepared water did not differ. Longevity of adults provided
with or without benzoic acid. water was signiÞcantly correlated with individual
Adult Life-History Traits. Mean adult emergence weight (r2 ⫽ 0.40; P ⱕ 0.0001).
rate (Table 3) differed signiÞcantly (F ⫽ 26.84; df ⫽ Egg Development and Clutch Size. Twenty virgin
3, 27; P ⱖ 0.0001) and ranged from 21.7 to 27.2% for the females per treatment, and four from the wild popu-
diet treatments, whereas 91.3% of the prepupae sam- lation, from 1 to 4 d postemergence were dissected and
pled from the wild yielded adults. Of those to emerge, examined for mature eggs, but none were found (Fig.
55.2Ð 60.5% were female (Table 3) with no signiÞcant 1 a and b). For each treatment, including the wild
difference across diet treatments or the wild popula- population, ⬍1% of the virgin females placed in plastic
tion (F ⫽ 1.05; df ⫽ 2, 22; P ⬎ 0.3660). Male black medicine cups laid eggs. None of these eggs hatched.
Thirty-two (72%) of the 45 mated-colony females that
Table 3. Percent survival of H. illucens larvae to the prepupal were dissected 2 d after mating (Fig. 1c) had com-
and adult stages and sex ratio for larvae reared on three diets and pletely formed eggs. Number of eggs per individual
for prepupae from a wild population
ranged from 323Ð 621, which accounted for ⬇13Ð26%
Larval Adult of their body weight. Six of the 10 mated females from
Sex ratio
Diet n survivorship emergence
(% female)
the colony that were held individually in 35-ml cups
to prepupae from prepupae deposited eggs. The number of eggs laid per individual
Gainesville diet 8 97.8 ⫾ 0.6A 27.2 ⫾ 6.7A 60.5 ⫾ 1.2A ranged from 206 to 639, which accounted for 7.9 Ð
CSMA 9 96.0 ⫾ 1.4A 23.9 ⫾ 4.0A 55.2 ⫾ 2.0A 23.4% of their body weight. Individuals dissected 3 d
Layer hen ration 9 96.1 ⫾ 0.4A 21.7 ⫾ 1.8A 58.0 ⫾ 3.4A after laying eggs (Fig. 1d) had no additional ovarian
Wild population 6 NA 91.3 ⫾ 4.8B 56.0 ⫾ 0.0A
development and little fat body present.
Means within a column followed by different letters differ signif- Egg Collection System. Mean eggs per ßute (F ⫽
icantly (P ⱕ 0.05; LSD, SAS Institute 1992). NA, not available. 0.72; df ⫽ 3, 71; P ⬎ 0.5456) and weight of eggs per ßute
May 2002 TOMBERLIN ET AL.: SELECTED LIFE-HISTORY TRAITS FOR BLACK SOLDIER FLIES 383

Table 4. Selected life-history traits of H. illucens reared on three diets and for adults reared from field-collected prepupae

Longevity, d adults Longevity, d adults Development, Development from

Diet treatment provided water not provided water d from egg egg to adult, d
Male Female Male Female to prepupa Male Female
Gainesville diet A9.3 ⫾ 0.4A,a A7.9 ⫾ 0.2A,a B6.0 ⫾ 0.2A,a B6.1 ⫾ 0.1A,a 22.5 ⫾ 0.7A 43.0 ⫾ 2.9A,a 43.4 ⫾ 2.1A,a
n ⫽ 106 n ⫽ 161 n ⫽ 159 n ⫽ 160 n⫽3 n⫽3 n⫽3
CSMA A9.7 ⫾ 0.4A,a A8.5 ⫾ 0.2A,a B5.9 ⫾ 0.2A,a B6.2 ⫾ 0.2A,a 23.4 ⫾ 0.3A 43.0 ⫾ 2.5A,a 43.0 ⫾ 1.3A,a
n ⫽ 103 n ⫽ 163 n ⫽ 145 n ⫽ 138 n⫽3 n⫽3 n⫽3
Layer hen ration A9.3 ⫾ 0.4A,a A8.5 ⫾ 0.3A,a B7.1 ⫾ 0.2B,a B6.4 ⫾ 0.2A,a 24.1 ⫾ 0.9A 40.4 ⫾ 2.4A,a 41.7 ⫾ 2.1A,a
n ⫽ 96 n ⫽ 151 n ⫽ 140 n ⫽ 138 n⫽3 n⫽3 n⫽3
Wild population A14.3 ⫾ 1.2B,a A14.2 ⫾ 0.9B,a B7.8 ⫾ 0.4B,a B8.2 ⫾ 0.4B,a NA NA NA
n ⫽ 25 n ⫽ 42 n ⫽ 31 n ⫽ 22

Means within a column followed by different capital letters differ signiÞcantly. Means for both sexes within a treatment with different lower
case letters differ signiÞcantly. Longevity means for a sex provided water and not provided water preceded by different capital letters differ
signiÞcantly (P ⱕ 0.05; LSD, SAS Institute 1992). NA, Not available.

(F ⫽ 0.83; df ⫽ 19, 20; P ⬎ 0.4798) were not signiÞ- larvae fed 5-d-old hen manure grew at half the rate of
cantly different among diet treatments, including the larvae fed 18 h old manure (D.C.S., unpublished data).
wild population (Table 5). Individual eggs ranged in Weight differences recorded for prepupae reared on
weight from 0.23 to 0.25 mg and mean number of eggs laying hen ration and those reared on the other diets
per ßute ranged from 603 for adults reared on Gaines- might be attributable to differences in calcium con-
ville diet and the wild population to 689 for adults tent (Table 1). Soldier ßy larvae sequester ingested
reared on layer hen ration. Mean weight of eggs per calcium and convert it to calcium carbonate, which is
ßute (clutch weight) ranged from 0.0145 g from adults secreted through the hypodermis and covers the larval
reared on the Gainesville diet to 0.0159 g for adults integument (Johannsen 1922).
reared on the CSMA larval ßy diet. Correlations be- Development time from egg to prepupae for the
tween mean egg weight and number of eggs per ßute black soldier ßy in our study was six to eight days
for adults reared on each diet and the wild population shorter than that recorded by May (1961) who re-
were signiÞcant (P ⬍ 0.05) (Table 5). The regression ported a mean development of 31 d under similar
analysis for mean egg number versus egg clutch weight temperature and humidity conditions. Differences be-
per ßute size was also signiÞcant for all treatments tween the two studies might be due to various factors,
including the wild population (Table 5). such as different larval densities (Morrison and King
The inability to determine a difference in the size of 1977) or food quality (Slansky and Scriber 1985, Dicke
egg clutches per diet treatment and the wild popula- et al. 1989). Furman et al. (1959) suggested these
tion was due to the early misconception that one ßute factors could delay black soldier ßy larval develop-
held the eggs of one female (Booth and Sheppard ment up to four months. However, May (1961) did not
1984). The number of eggs deposited per ßute was provide detail sufÞcient for comparison.
limited by ßute size. The number of eggs deposited in Differences in percent emergence per diet treat-
the smaller ßute (1 by 3 mm) averaged 431 eggs per ment and those collected from the wild might also be
ßute, whereas those in the larger ßute (2 by 5 mm) due to larval diet and being held in a sub-optimal
averaged 1,157 eggs per ßute. Based on measurements habitat. Larvae in our experiment were held and fed
of the depths that eggs were deposited in the different in containers that resulted in continuous contact with
sized ßutes, individuals were able to insert their ab- their excreta. Sheppard (1983) determined that wild
domen and ovipositor ⬇1 cm into the larger ßute and soldier ßy larvae remain on the surface of the wastes
⬇0.7 cm in the smallest ßute. Egg size was not signif- accumulating below layer facilities and feed on fresh
icantly different across diet or ßute treatments (F ⫽ manure as it is produced. Manure digested by the
0.83; df ⫽ 3, 71; P ⬎ 0.4798). larvae then accumulates below the larvae. Because the
soldier ßy larvae in our experiment were held in closed
containers, they fed on media that was mixed with diet
Discussion
residue and their own waste. Therefore, the soldier ßy
All three diets are suitable for rearing the black larvae might have been forced to feed to a degree on
soldier ßy. They are nutritionally similar, which prob- their own wastes, which we suspect they normally
ably explains why Þnal larval weight and caloric con- avoid. Larvae of other dipteran species, such as the
tent of diet-reared prepupae were comparable. How- house ßy, need fresh manure daily for optimal growth
ever, lab-reared prepupal weight differed from the (Morgan and Eby 1975). This need for fresh manure
wild prepupae sampled, which might be due to wild is because anaerobic organisms digest manure as it
larvae having access to additional nutrients (Slansky ages, resulting in fewer nutrients for the larvae and
and Scriber 1985, Dicke et al. 1989), such as the con- thereby suppressing larval growth (Beard and Sand
tinuous stream of fresh manure in poultry facilities, 1973). Developmental time from egg to the adult did
not mixed with old resources as in our tests. Recent not differ between sexes for wild black soldier ßies, as
laboratory studies determined that black soldier ßy well as those reared on diet in the laboratory.
384 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 95, no. 3

Fig. 1. Ventral view of the dissected abdomen of a female black soldier ßy. (a) Freshly emerged (⬍24 h) with fat body
present; (b) freshly emerged (⬍24 h) with fat body removed; (c) 2 d after mating remaining fat body intact; (d) 3 d after
laying eggs with remaining fat body intact.

Water, unlike food, is essential for adult black sol- adults being unable to mate effectively and therefore
dier ßies to reproduce (Sheppard et al. 2002), which not laying viable eggs. We determined that individuals
might be due to the less vigorous and dehydrated reared on the diets and provided water lived 1Ð2 d
May 2002 TOMBERLIN ET AL.: SELECTED LIFE-HISTORY TRAITS FOR BLACK SOLDIER FLIES 385

Table 5. Comparison of clutch weight (grams), and egg weight (milligrams) for H. illucens reared on three diets and for field-collected
prepupae and the correlation and regression analyses for clutch weight (milligrams) and eggs per clutch per treatment

Treatment Correlationa Regression models for eggs per clutch Clutch weight Egg weight
2
Gainesville diet r ⫽ 0.88, df ⫽ 20 ⫽ 173.87 ⫹ 29,372 (clutch weight) 0.0145 ⫾ 0.00012A 0.24 ⫾ 0.012A
n ⫽ 20 F ⫽ 62.8; df ⫽ 1, 18; P ⬍ 0.0001
CSMA r2 ⫽ 0.90, df ⫽ 15 ⫽ 143.15 ⫹ 31935 (clutch weight) 0.0159 ⫾ 0.00018A 0.25 ⫾ 0.011A
n ⫽ 15 F ⫽ 55.5; df ⫽ 1, 13; P ⬍ 0.0001
2
Layer ration feed r ⫽ 0.93, df ⫽ 20 ⫽ 124.64 ⫹ 35346 (clutch weight) 0.0158 ⫾ 0.00012A 0.23 ⫾ 0.008A
n ⫽ 20 F ⫽ 122.3; df ⫽ 1, 18; P ⬍ 0.0001
2
Wild population r ⫽ 0.98, df ⫽ 20 ⫽ 25.12 ⫹ 38887 (clutch weight) 0.0153 ⫾ 0.00012A 0.25 ⫾ 0.004A
n ⫽ 20 F ⫽ 686.8; df ⫽ 1, 18; P ⬍ 0.0001

a
PearsonÕs correlation, signiÞcance set at P ⱕ 0.05 (SAS Institute 1992).
Means within a column followed by different letters differ signiÞcantly.

longer than those not provided water. In contrast, wild estimate of the total number of eggs produced by a
adult soldier ßies provided water lived 4 d longer than female. Additionally, oviposited eggs might not be as
adults reared on the diets. This signiÞcant difference accurate due to variables inßuencing number of eggs
in longevity might be explained by caloric content of produced, such as time from emergence to mating.
the prepupae. Prepupae reared on the diet treatments Rogers and Marti (1994) determined that the age of
contained ⬇365 cal, whereas the wild prepupae con- female fall armyworms, Spodoptera frugiperda (J. E.
tained 1,309 cal per prepupa. The additional calories Smith) (Lepidoptera: Noctuidae), at Þrst mating sig-
provided energy that might have resulted in increased niÞcantly inßuences reproductive potential and that
longevity of the wild adults. extended female virginity resulted in reduced fecun-
Egg development studies were problematic because dity and fertility.
of misconceptions concerning the deÞnition of a black
The black soldier ßies in our studies reproduced
soldier ßy egg clutch, or eggs oviposited by one indi-
without feeding and apparently relied on fat reserves
vidual. Booth and Sheppard (1984) stated that the
average number of black soldier ßy eggs per clutch was acquired during larval development for adult sur-
998 ⫾ 78 (mean ⫾ SE). They collected eggs using the vival and egg production (D.C.S., unpublished data).
openings in corrugated cardboard (ßutes) as the ovi- Therefore, any delay in mating could result in re-
position site. Before Booth and Sheppard (1984), re- sources being reallocated from producing eggs to pro-
searchers reported black soldier ßy egg clutches with longing the adult stage. Similar to our hypothesis for
119 Ð502 eggs (Gonzalez et al. 1963) and 205Ð 802 (Ste- the black soldier ßy, Chippindale et al. (1993) deter-
phens 1975) in dried hen manure and on bananas, mined that food quality was directly related with egg
respectively, which are similar to our eggs per clutch production and inversely related to adult longevity for
data determined for individuals dissected 2 d after Drosophila melanogaster Meigen (Diptera: Drosophi-
mating. lidae). Higher quality food resulted in greater egg
Methods described by Booth and Sheppard (1984) production, but reduced longevity in fruit ßies. Ad-
for collecting black soldier ßy eggs are suitable. How- ditionally, the number of times a female arthropod
ever, these methods are not recommended for esti- mates can inßuence the number of eggs deposited. As
mating clutch size for individual soldier ßies. Egg suggested by Foster and Ayers (1995) for female
clutches from diet-reared and wild adults were smaller Epiphyas postvittana (Walker) (Lepidoptera: Tortri-
than those recorded by Booth and Sheppard (1984), cidae), mating more than once might be attributed to
but individual egg size was similar. The reduction in male deÞciencies, but might result in greater numbers
egg number per clutch for our experiment and that of eggs being fertilized and deposited.
recorded by Booth and Sheppard (1984) is most likely Most of the literature on the black soldier ßy is
due to ßute size, which we determined can limit the restricted to its use as a biological control agent of
number of eggs deposited per ßute. Therefore, our
house ßies and as a manure management agent in
results really estimate number of eggs per ßute, which
poultry facilities. Our study provides additional infor-
also might have been the case for Booth and Sheppard
(1984). However, the ßute size they used is unknown. mation on the life history of this species reared on
We suspect that multiple individuals could deposit three diets, as well as for those collected from the wild.
eggs in a single ßute, which would skew clutch size Such information is necessary for developing its po-
estimates. Egg clutch weight varied ⬍1% when allow- tential as a biological control and waste management
ing individuals to oviposit in cardboard ßutes of equal agent in livestock and poultry production. The three
size. However, egg clutches oviposited by mated in- diets examined in this study are suitable for rearing
dividuals collected from the black soldier ßy colony black soldier ßies. However, because of the large dif-
and placed in 35-ml cups and those dissected 2 d after ferences between wild and laboratory-reared speci-
mating varied as much as 75%. We hypothesize dis- mens, further research is needed to reÞne and improve
secting females 2 d after mating provides an accurate larval rearing for mass production of this species.
386 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 95, no. 3

Acknowledgments Johannsen, O. A. 1922. Stratiomyid larvae and puparia of the

North Eastern States. J. N.Y. Entomol. Soc. 30: 141Ð153.
We are grateful to G. Herzog for allowing us the use of an May, B. M. 1961. The occurrence in New Zealand and the
environmental chamber and J. Dutcher for the use of the life-history of the soldier ßy Hermetia illucens (L.)
bomb calorimeter during our experiments. We also thank J.
(Diptera: Stratiomyidae). N.Z. J. Sci. 4: 55Ð 65.
Ruberson, W. Berisford, R. Noblet, P. Adler, N. Hinkle, J.
Morgan, N. O., and H. J. Eby. 1975. Fly protein production
Butler, D. Cochran, and two anonymous reviewers for valu-
from mechanically mixed animal waste. Israel J. Entomol.
able suggestions on earlier versions of the manuscript.
10: 73Ð 81.
Morrison, R. K., and E. G. King. 1977. Mass production of
References Cited natural enemies, In R. L. Ridgway and S. B. Vinson [eds.],
Biological control by augmentation of natural enemies.
Axtell, R. C., and J. J. Arends. 1990. Ecology and manage- Plenum, New York
ment of arthropod pests of poultry. Annu. Rev. Entomol. Newton, G. L., C. V. Booram, R. W. Barker, and O. M. Hale.
35: 101Ð126. 1977. Dried Hermetia illucens larvae meal as a supple-
Beard, R. L., and D. S. Sand. 1973. Factors affecting degra- ment for swine. J. Anim. Sci. 44: 395Ð 400.
dation of poultry manure by ßies. Environ. Entomol. 2: Rogers, C. E., and O. G. Marti, Jr. 1994. Effects of age at Þrst
801Ð806.
mating on the reproductive potential of the fall army-
Bondari, K., and D. C. Sheppard. 1981. Soldier ßy larvae as
worm (Lepidoptera: Noctuidae). Environ. Entomol. 23:
feed in commercial Þsh production. Aquaculture 24: 103Ð
322Ð325.
109.
SAS Institute. 1992. SAS/STAT userÕs guide, version 6, 4th
Booth, D. C., and D. C. Sheppard. 1984. Oviposition of the
black soldier ßy, Hermetia illucens (Diptera: Stratiomyi- ed. SAS Institute, Cary, NC.
dae): eggs, masses, timing and site characteristics. Envi- Sheppard, D. C. 1983. House ßy and lesser house ßy control
ron. Entomol. 13: 421Ð423. utilizing the black soldier ßy in manure management
Chippindale, A. K., A. M. Leroi, S. B. Kim, and M. R. Rose. systems for caged laying hens. Environ. Entomol. 12:
1993. Phenotypic plasticity and selection in Drosophila 1439 Ð1242.
life history evolution. I. Nutrition and the cost of repro- Sheppard, D. C., G. L. Newton, and S. A. Thompson. 1994.
duction. J. Evol. Biol. 6: 171Ð193. A value added manure management system using the
Dicke, M., M. DeJong, M.P.T. Alers, F.C.T. Stelder, R. Wun- black soldier ßy. Bioresource Tech. 50: 275Ð279.
derink, and J. Post. 1989. Quality control of mass-reared Sheppard, D. C., J. K. Tomberlin, J. A. Joyce, B. C. Kiser, and
arthropods: nutritional effects on performance of pred- S. M. Sumner. 2002. Rearing methods for the black sol-
atory mites. J. Appl. Entomol. 107: 462Ð475. dier ßy, Hermetia illucens (L.) (Diptera: Stratiomyidae).
Dutcher, J. D. 1983. Ecological energetics and pecan weevil J. Med. Entomol.
bionomics. J. Ga. Entomol. Soc. 18: 208Ð213. Slansky, F., Jr., and J. M. Scriber. 1985. Food consumption
Foster, S. P., and R. H. Ayers. 1995. Multiple mating and its and utilization. In: G. A. Kerkut and L. I. Gilbert [eds.],
effects in the light brown apple moth, Epiphyas postvit- Comprehensive insect physiology, biochemistry and
tana (Walker). J. Insect Physiol. 42: 657Ð667. pharmacology, vol. 4. Pergamon, Oxford, England.
Furman, D. P., R. D. Young, and E. P. Catts. 1959. Hermetia Stephens, C. S. 1975. Hermetia illucens (Diptera: Stratiomyi-
illucens (Linnaeus) as a factor in the natural control of dae) as a banana pest in Panama. Trop. Agric. (Trinidad)
Musca domestica Linnaeus. J. Econ. Entomol. 52: 917Ð921. 52: 173Ð178.
Gonzalez, J. V., W. R. Young, and M. R. Genel. 1963. Re- Tingle, F. C., E. R. Mitchell, and W. W. Copeland. 1975. The
duccion de la poblacion de mosca domestica en gallinaza soldier ßy, Hermetia illucens, in poultry houses in North
por la mosca soldado en el tropico. Agric. Tec. Mex. 2: Central Florida. J. Ga. Entomol. Soc. 10: 179 Ð183.
53Ð57. Tomberlin, J. K. 2001. Biological, behavioral, and toxicology
Hogsette, J. A. 1985. New diets for production of house ßies studies on the black soldier ßy (Diptera: Stratiomyidae).
and stable ßies (Diptera: Muscidae) in the laboratory. J. Ph.D. dissertation, University of Georgia, Athens.
Econ. Entomol. 85: 2291Ð2294.
James, M. T. 1935. The genus Hermetia in the United States
(Diptera: Stratiomyidae). Bull. Brooklyn Entomol. Soc. Received for publication 15 October 2001; accepted 11 Feb-
30: 165Ð170. ruary 2002.