Ecolocv eNo Bnnevron Model for Use in Mass-Production of Acheta domesticus (Orthoptera: Gryllidae) as Food

MEGHA N. PARAJULEE, GENE R. DTFOLIART, ENO DAVID B. HOGG Department of Entomology, University of Wisconsin, Madison, WI 53706

EiSt"iilx"i;'"'Jil;L1itl111l]3?11., ABSTRAcT e p,od,,"uJ,, .n" harvest or a qrede- termined numbei of eggs of house cricket, Acheta domesticus (L.), per day by regulating the numbers and ages of adults in the breeding colony. With a discard age of 24 d, the production model predicted a sustainable harvest of -4,000 (4,440) and 6,000 (6,660) crickets per day when four or six pairs, respectively, of newly emerged adults were added per day to an initial breeding colony of 50 pairs. Natality was,based on the number of nymphs surviving to 7 d per surviving female, after which little nymphal mortality oc- curted. Orripositional surface area availability was not a limiting factor in egg production.

KEY WORDS Acheta domesticus, human food, mass-production

Tnn nousr cRICKET, Acheta domesticus (L.), is are harvested on a continuing basis. The system a cosmopolitan, omnivorous insect that is easily is expandable depending upon the daily produc- reared under conffned conditions (Patton 1978). tion desired. Previous research (Jordan & Baker 1956; Patton Previously, it has been difficult to determine 1963, 1967; Polt l97l; Clifford et al. 1977; Mc- proper number of eggs to seed per cage without Farlane 1985; Parajulee & DeFoliart 1993) de- sorting and counting thousands of eggs from the scribed the rearing methods and emphasized the oviposition medium.each day. The two objec- widespread use and suitability of house crickets tives of our study were to (1) develop and test a for laboratory studies. In the United States and production model based on age-specific natality Canada it is widely used as ffsh bait and for and cumulative net maternity of crickets that maintaining insectivorous vertebrates in captiv- would accurately predict the number and ages of ity. Finke et al. (1989) found that the house crickets needed in the breeding colony to pro- cricket is of high protein quality and slightly duce the desired number ofeggs per day, and (2) superior to soy protein at all levels of feeding to determine whether the area of ovipositional when fed to weanling rats, and Nakagaki et al. surface available was limiting to daily egg pro- (1987) found that dried house crickets are a duction. source ofhigh quality protein for broiler chicks. The cricket is recommended by Taylor & Carter Materials and Methods (1976) as an ingredient in gourmet recipes. Na- kagaki & DeFoliart (1991) have shown the Developing the Production Model. To deter- cricket to be competitive compared with other mine cricket natality and cumulative net ma- livestock on the basis of food conversion effi- ternity, we used net reproductive rate (Rn) as ciency. Because of its palatability, protein qual- defined as2,l,m, (Andrewartha & Birch 1954), ity, and high food conversion efficiency, the where survivorship (1") is the probability of an cricket is a promising candidate for wider promo- individual female attaining age r (days) and na- tion as food and animal feed (DeFoliart 1989). tality (m,) is the number of female nymphs pro- We designed a mass-rearing system which in- duced by a female between ages r-l and r. We volves a breeding colony and 32 rearing units, define cumulative net maternity as 2(Ro) to in- each unit providing enough space (50 by 44by clude male offspring. 20.5 cm) to rear =6,000 crickets to harvestable Three simultaneous replicates of 50 pairs of size (eighth instar). One cage is seeded each day newly emerged adult crickets were confined in with enough eggs to allow for mortality, and the glass terraria (50 bv 30 by 26 cm) covered with a 6,000-cricket cohort is harvested 32 d later. Fol- perforated aluminium lid to prevent crickets lowing harvest, each cage is reseeded (one cage from escaping. We started the experiments with per day) with 9-d-old eggs (ready to hatch) from 50 pairs because ofaberrations in cricket behav- the breeding colony. Thus, 6,000 crickets per day ior at low population densities (McFarlane

0022-0493 193I 1 424 -1428$02.00 l0 O I 993 Entomolo gical S ociety of America October 1993 Penelur,rn ET AL.: PnooucrloN Monnr, ronAcheta domesticus t425

1962). The crickets were fed the Animal Nutri- adults per day in an initial breeding colony of50 tion Research Council Reference Chick diet (Na- pairs. tional Research Council 1977) for broiler-type Testing the Production Model. The initial chicks, modified by the addition of 0.5% NaCl breeding colony consisted of 50 pairs of newly and 3Vo fish protein (McFarlane 1964). Observa- emerged adult crickets conffned in an aluminum tions were made daily to determine survival of cage (50 by 44by 20.5 cm), to which four pairs of females and to remove dead crickets. Feed and newly emerged adults were added daily (herein- water were provided ad libitum. The feed was after referred to as a four-pair colony). Petri provided in a petri dish at one end ofthe terrar- dishes (I5 cm diameter) were used as egg trays ium. Fresh tap water was supplied to the breed- and moistened peat moss was provided as an ing colony using a standard chick watering de- ovipositional substrate as described above. For vice consisting of a 0.9-liter glass jar with a 50 d, individual egg trays were exposed to the screw-on plastic pan base. Aluminum screening breeding colony for a 24-h period, allowing for under the water outlet prevented the crickets oviposition, then removed and incubated at 34 -r from crawling into the opening and drowning. 2oC under continuous light. The egg trays were The water bottles were emptied, washed, and placed in vented storage boxes and nymphs were refilled every 3 d to prevent crickets from drink- reared, killed, and counted as described previ- ing fouled water. Three pieces of cardboard ously. chicken egg cartons (30 by f5 cm) were placed in In a second experiment under the conditions each terrarium to provide resting and hiding sur- and procedures just described, six pairs of newly faces for the crickets. emerged adults were added daily to an initial Petri dishes (15 cm diameter) were used as egg breeding colony of 50 pairs (hereafter referred to trays and filled with moistened peat moss as an as a six-pair colony). Eggs were harvested and ovipositional substrate. For 45 d, individual egg nymphs reared for 7 d, then killed by freezing trays were exposed to the crickets for 24 h-peri- and counted as before. ods, allowing for oviposition. The trays were In both experiments, older adults were not re- then removed, covered with a lid, and incubated moved from the breeding colony, but dead adults at 34 -r 2"C (Roe et al. 1980). Just before egg were removed when found. hatch, the trays were placed in vented plastic Oviposition Substrate Area as a Limiting Fac- storage boxes (34 by 2I by 9 cm) to rear the tor. This experiment was conducted as described nymphs. Condensed water on the undersurface for the four-pair colony, except that the oviposi- of the lid was removed by paper towels to pre- tional surface area was reduced by 50Vo. A semi- vent drowning of newly eclosed nymphs. After circular wooden block of 5 cm thickness, -2 the eggs hatched (incubation period of 9 d at times the height of the petri dish rim, was fftted 34 -r 2'C), the lid was removed from the egg tray against the lS-cm-diameter petri dish to provide and food and water were provided for the newly a barrier against oviposition in that half of the eclosed nymphs. A cotton dental wick (I by 5 cm) petri dish. The other half was filled with moist- was inserted into a vial containing water and laid ened peat moss. Eggs were harvested after ex- on its side on the storage box floor to provide posing the modified egg trays to the breeding water for the nymphs. One or two folded paper colony for a24-h period for 46 consecutive days. towels were provided in each storage box to in- Egg trays were incubated and the nymphs were crease surface area and facilitate cricket molting. reared for 7 d before freezing and counting as Finely ground feed was sprinkled over the paper before. The mean production from this experi- towels. Nymphs were reared for 7 d, then killed ment was compared with that of the production by freezing and counted. Patton (1978) deter- achieved from the four-pair colony experiment mined that at 32 I l"C, A. domesticus molts by using t test statistics. occur at :3-d intervals, thus at34 -+ 2'C, most of instar our nymphs were in the third when killed Results and Discussion for counting. Clifford et al. (1977) found that most A. domesticas mortality occurs in the first and Developing and Testing the Production second instars with essentially l00Vo survival Model. Fig. I shows the age-specific natality thereafter; thus, the number of crickets surviving (mean number of offspring surviving for 7 d per to 7 d should closely approximate the total num- surviving female) for the cricket cohorts. A peak ber of crickets surviving to harvest. Our fecun- of 90 offspring per female occurred on day ll, dity model, then, is based on the number of which repres ents 6.4Vo of cumulative net mater- nymphs surviving to the seventh day. nity, and declined to fewer than 3 offspring per The life-history statistics, such as net repro- female by day 45 (Fie. l). No adult mortality was ductive rate and cumulative net maternity, were recorded before age 24 d, and survivorship de- calculated from the age-speciffc life-table analy- clined to TIVo by the termination of the experi- sis. These life-history statistics were then used to ment on day 45 (Fig. 2). Although the survivor- examine the dynamics of production for the ad- ship seems apparently high even on day 45 dition of either 4 or 6 pairs of newly emerged (70Vo), reduced competitive ability of older fe- L426 JounNer- or EcoNol,rrc ENToMoLocy Vol. 86, no. 5

100 A 80 o o U' =E 60 F o- ul U' Y l! o lt c o o zct o z -.- obssrued Predicted

0 10 20 30 - 40 50 50

FEMALE AGE (DAYS) Fig. 1. Number of A. domesticus offspring (surviv- B ing for 7 d) produced per adult female per day at 34 + 2'C. Each data point represents the mean of three co- horts. o F 6000 males contributes to very little effect on overall vLrJ population growth (see concept o below for the of E 4ooo discard age and the poor competitive ability of o the older females). The mean cumulative net ma- 2ooo ternity for the generation (twice R") was 1,407. 9 + Observed Using the natality (Fig. 1) and survivorship Predicted (Fig. 2) data from the single-aged cricket cohorts, - 5040 60 we developed a model to predict the harvest for 2010 30 a mass-rearing production system: DAYS

Et+t:2(uuwt.) Fig. 3. Daily observed and predicted production of i 7-d-old A. domesticus offspring from a breeding colony of 50 pairs with four pairs (A) and six pairs (B) of newly No,t = P emerged crickets added daily.

Nr,r: sr - rNl - j,t - l,i>O from a single-aged colony in that a fixed number ofnewly eclosed adult pairs are added each day, where E represents the number of eggs, N is the resulting in a population with mixed ages. How- number of adult females, i is adult age class, f is ever, once the adult age structure in such a col- time, b is natality (male plus female offspring), ony becomes artiffcially "stable" (i.e., all age p is the number of cricket pairs added daily, and classes are represented with numbers relative to s is survivorship. A production colony differs their survival expectations), the daily production of offspring should be sustainable and equal to a multiple of the cumulative net maternity (i.e., 1.00 1,407 per cricket pair added). Thus our model would predict a harvest of 5,628 and 8,442 nymphs per day from colonies in which four o- 0.90 pairs and six pairs, respectively, were added o! daily. E o The production-system results (Fig. 3) dis- 0.80 agreed with our model predictions in that for Et both the four-pair and six-pair colonies, the ac- o tual sustainable cricket production was substan- 0.70 tially less than predicted. These differences were probably caused by reduced competitive abilities of the older females in the mixed-age 0.60 0 '1 0 20 30 40 50 colonies, In the presence of younger females, older females tended to spend less time around FEMALE AGE (DAYS) the egg trays and more time in the egg carton Fig. 2, Survivorship of female A. domesticus adrlts resting places (M.N.P., unpublished data). The at34 ! 2'C. older females could be recognized by their dark, October 1993 Pena;ur,nr ET AL.: PnooucrroN Moner- ron Acheta domesticus r427

rough, dirty wings compared with lighter, shiny, and cleaner wings of young females. However, it did not appear that older females experienced an increased mortality rate in the mass-rearing col- 3 6000 D compared with the single-aged colonies. F onies rll Carey & Vargas (1985) found that in mass rear- Y g 4000 ing three species of tephritids, eggs laid after E 3 wk of adult life contributed little to population o growth. To compensate, they invoked an artiff- cially imposed last day of reproduction, termed 2 2ooo + HALFSPACE + the adult discard age. Under our system of daily FULLSPACE introducing new cricket adults into the breeding colony, the data from the four-pair colony show a 0 10 20 30 40 50 60 sharp drop in cricket production on day 24 (FiS. DAYS 3A), indicating a reduction in oviposition by the Effect ofreducing the oviposition substrate females in the original 5O-pair cohort and sug' Fig.4. area by half on the daily production of 7-d-old A. do- gesting an adult discard age of 24 d. It would not mesticus offspring from a colony of 50 pairs with 4 pairs il be practical in our case to actually remove crick- of newly emerged adults added daily. ets from the colonies. However, assuming that reproduction in a mixed-age colony becomes negligible after age 24 d, we can impose an ef- When the oviposition substrate area was re- fective discard age, which results in a revised duced, the increased density ofcrickets resulted previ- cumuiative net maternity value of I,ll0 offspring in loss of both oviposition substrate and substrate out of per female. With this discard age, our model thus ously laid eggs by adults kicking (M.N.P., unpublished data). When a predicts a harvest of 4,440 and 6,660 nymphs per the trays day for the four-pair and six-pair colonies, re- larger surface area was available, there also was the substrate, but it tended to occur spectively. In both cases, the experimental re- kicking of near the center of the dish, resulting in less sub- sults were in general agreement with these pre- strate being kicked out of the dish. The loss of dictions. Observed mean daily production was oviposition substrate is negligible using a lS-cm- 3,995 (t861) and 6,535 (t688) nymphs for the diameter petri dish fflled to only three-fourths of respectively, from four-pair and six-pair colonies, its depth with moistened peat moss. Dry peat days 25 to 50 of the experiments. moss is light in weight, which increases the pos- production experiment did not One issue our sibility of loss from disturbance. address directly is the sustainability of yields beyond 50 d. We assume that the daily yields we observed could be maintained indefinitely. Acknowledgments However, although reproduction by older fe- We thank Sharmila Parajulee (University of males appears to be reduced in mixed-age colo- Wisconsin-Madison) for technical assistance, and nies, survivorship (Fig. 2) may not be similarly James Carey (University of California, Davis) and two affected. It is possible that over time, old crickets anonymous reviewers for their helpful comments. This could build up to levels at which they would research was supported in part by the Research Divi- interfere with colony productivity, but this is un- sion, College of Agricultural and Life Sciences, Uni- (Hatch likely because old crickets and the young, pro- versity of Wisconsin-Madison Project 3038). ductive crickets appear to compete only for the essentially unlimited number of resting places References Cited provided by the egg cartons. Area Limiting Andrewartha, H. G. & L. C. Birch. 1954. The distri- Oviposition Substrate as a bution and abundance of animals. University of Factor. We detected no density-dependent de- Chicago Press, Chicago. pression in per-cricket production between the Carey,J. R. & R, I. Vargas. 1985. Demographic anal- four-pair and six-pair cages. The area ofoviposi- ysis of insect mass rearing: a case study of three tion substrate rather than overall cage volume tephritids. J. Econ. Entomol. 78:523-527. could be the limiting factor for cricket mass- Clifford, C. W., R. M. Roe & J. P.Woodring. 1977. production, and increasing only this area might Rearing methods for obtaining house crickets, increase the adult cricket density in a cage be- Acheta domesticus, of known age, sex, and instar. yond the recommended levels (Patton 1978). Our Ann. Entomol. Soc. Am. 70 69-74. DeFoliart, G. R. 1989. The human use of insects as results, however, indicated that the oviposition feed. Bull. Entomol. Soc. Am. (Fig. food and as animal substrate area was not limiting a). Nymph 35:2%-35. -f + production was 3,737 1,555 (mean SD) when Finke, M. D., G. R, DeFoliart & N. J. Benevenga. the ovipositional surface was reduced by half, 1989. Use of a four-parameter logistic model to and 3,773 -+L,459 when the entire area of the evaluate the quality ofthe protein from three insect petri dish was available. species when fed to rats. J. Nutr. 119: 864-871. 1428 JOURNAL oF EcoNoMrc ENToMoLocY Vol. 86, no. 5

Jordan, B. M. & J. R. Baker. f956. The house cricket Parajulee, M. N. & G. R. DeFoliart. 1993. A simple (Acheta domestica) as a laboratory animal. Ento- method of heat supplementation for cricket mass- rearing in the absence of controlled-temperafure mologist 89: 126-127. j. McFarlane, E. 1962. A comparison of the growth equipment in developing countries. Inst. Agric' J, (in the house cricket (Orthoptera: Gryllidae) reared Anim. Sci. press). of Rearing the house cricket, singly and in groups. Can.J.Zool.40: 559-560. Patton, R. L. 1963. commercial feed' Ann. protein requirements the house Acheta domesticus, on 1964. The of 250-257' 645- Entomol. Soc. Am. 56 cricket, Acheta donxesticus L. Can. J. Zool. 42: 1967. Oligidic diets for Acheta domesticus (Or- 647. thoptera: Gryllidae). Ann. Entomol. Soc. Am. 60: 1985. Acheta domesticus,pp.42T-434' tn P. Singh 1238-t242. & R. F. Moore [eds.], Handbook of Insect Rearing, 1978. Growth and development parameters for vol. I. Elsevier, Amsterdam. Acheta domesticus. Ann. Entomol. Soc. Am. 7I: 40- Nakagaki, B. J. & G. R. DeFoliart. 1991. Compar- 42. ison of diets for mass-rearing Acheta domesticus Polt, J. M. f971. Experiments in animal behavior. (Orthoptera: Gryllidae) as a novelty food, and com- Am. Biol. Teach.38: 475-477. parison of food conversion efficiency with values Roe, R. M., C. W. Clifford & J. P.Woodring. 1980. reported for livestock. J. Econ. Entomol. 84: 891- The effect of temperature on feeding, growth, and 896. metabolism during the last larval stadium of the Nakagaki,8.J., M.L. Sunde & G. R. DeFoliart. 1987. female house cricket, Acheta domesticus. J. Insect Protein quality of the house cricket, Acheta domes- Physiol. 26:639-644. ticus, when fed to broiler chicks. Poult. Sci. 66: Taylor, R. L. & B. I. Carter. 1976. Entertaining with 1367-r371. insects, or: the original guide to insect cookery. National Research Council. 1977. Nutrient require- Woodbridge, Santa Barbara, CA. ments of domestic animals. No. 1: nutrient require- ments of poultry. National Academy of Sciences, Receioed for publication 7 September'7992; ac' Washington, DC. cepted 3 Maa 1993.