genetics of endosulfan resistance in hypothenemus hampei...

INSECTICIDE RESISTANCE AND RESISTANCE MANACEMENT

Genetics of Endosulfan Resistance inpHypothenemus hampeì e

Sex Determination (Coleoptera: Scolytidae): Implications for Mode of !

LUC O. BRUN, D. MAXM'ELL SUCKLING,' RICHARD T. ROUSHI2 VERONIQUE GAUDICHON, HAIGANOUSH PREISLER," ANI)

JACQUELINE L. HOBERTSON'3

Institut Franpis de Recherclie Scientifirpr polir le D6vPlop elnent en C'oopérotion (OIISTOM), B.P. As, NoilniCa, New CaleIonia

ABSTRACT The genetics of endosulfiui resistnnce in Hyi"'tl¿cneitirls h t r i t i p c i (Ferrari) w a s studied through clussical crosses to deterinine the degree of do~n in~u~ce and ni~lul)er CJf genes involved. After adult beetles were spr~iyed, Inortdity was recorded at 6 11 iinrl 7 d. Tests of the males froni reciprociil FI crosses intlicatetl that resistance is ses-linket1 or tliut piterIial diromosonies are inactive i n the sons. Responses of FI feni&s nfter 6 11 iiltlicnted degrees of domiiinnce of -0.38 -C o.on und -0.25 k 0.0:3 for RI< x s and ss x 13 crosses, respectively. In contrast, dolninance after 7 d WIS -0.17 a 0.02 ant1 -0.02 -C 0.02. tipp'irently intlicatilig ii trend towiirtl cotlominance over tinir. Hesponses of btickcrosses of the FI geiieration t(J I)otli parental lines and of FZ progeny were inconsistent with results predicted udien ossiinii~~g siinple bleritleli~in inlieritance. These results illiistrnte the practical ut&iesses of buckcrosses il1 estimating the nilinber of genes that control resistance in sitiiutiolls where Lioussaiys ilre inconsistent or where discrimination Iie+sveeii genotypes is p o r .

Hyptheneiiirrs ~ ¿ m z p e i (FEHRAHI) is a major pest of coffee worldwide and is higlily resistant to endosulfan in New Caledonia (Brun et al. 1989, 1992). Resistance has been confined to five regions on the East Coast, where the existence of resistance is significantly more frequent in the newer sunny plantations that liad been treated witli endosulfan during the preceding 12 mo than in fields that liad not been treated recently or in older fields under native forest canopy (Brun et al. 1990). Par- kin et al. (1992) related variation in resistance levels within fields reported in Brun & Suckling (1992) to operational fwtors s~icli as application from truck-mounted sprayers and tlie type of field. Cessation of endosulfan use led to a reduction in resistance frequency, whereas continued use in- creased the resistunce frequency (Brun & Suckling 1992).

Although much lias been learned about the evo- lution of resistance in H. haiiapei through monitor- ing and without any understanding of tlie under- . lying genetics, several other important factors, such as the effective doniinance of resistance

The Horticulture arid Foc~tl KesrRrcli Iiastitiatc of New Zea- land, P.O. Box 51, Lincolri, Oiii terl~ury, New Zealmid.

sity, Itlanca, NY 1485:3.

USDA Forest Servkr, P.O. ß o u 24.5, 13crkvley CA H4701.

' D ~ I ) X ~ I I W I I ~ Of E ! i t ( J ~ i ( ~ h ~ . C(Jli1St~Jck I-I;d\, C<Jrllt.\l Univer-

Pacific SlJrlttawest Forest aiid l<ulige Experiinent Station.

(Roiish & McKenzie 198í, Roiish & Daly 1990), cannot be studied withoiit deterinining tlie inode of inheritance of the trait. Thus, we sought to de- termine the degree of dominance and mode of inheritance of endosulfh~~ resistance in H. hmr~pci by using backcross nietliocls.

Materials and Rlethods

Insect Strains. Susceptible (S) and resistant (R) strains of H. hnmpei were collected in July 1990 froni coffee berries at the roadsides of fields near La Foa (LA 2, an S strain identified in Bnin et al. 19S9) and Ponérihouen (designated PN106). Fe- ninles of the K strain were mated, selected with 10,000 ppni of endosulfan (as described below) for four generations, with ZOO-l,OOO females selected per generation. Tliis concentration was used because it exceeded the LCBM9 (2,500 ppni) of the offspring of beetles collected directly froni the field. Survivors were reared, mated and selected, and used as the parent R strain. Lan-ae were reared on tlie artificial diet described by Brun et al. (1993).

Direct-Spray Teclinique. Concentr t' ion-mortality responses of R and S H. haiiipei with the direct spray technique after 6 11 have been reported previously (Brun et al. 1989, 1990). For exam- ination of tlie genetics of resistance, we recorded

nus hampeì ode of

3oiisli fi- Daly 1990). etermining the niocle iiis, we sought to de- aiice and mode of in- aiice in H. haiiipei by

Ht:tl1ods

: (S) and resistant (R) ollected in July 1990 atlsides of fields near -ntified iii Brun et al.

Concen tration-mor- I f . h i t p i with the

5 I I Jiave been report- 89, 1990). For exain- (sistance, we recorded

tlie survivors after G h and 'i d. A glass ring was used to confine adults un a piece of filter paper during spray application. Each test of females con- sisted of 2-S replicates (generally 90-120 beetles, \vitIl a range of 50-480) per concentration und genotype of the females. Highly female-biased sex ratio prevented extensive testing of males, which was limited to samples of 22-82 (4-9 replicates of 5-10 indi\iduals dependent on emergence). A Pot- ter spray tower (Potter 1952), calibrated to deliver 1.6 nig/cin', was used to apply 2 nil of licpiid onto each ,pmple. Adults were held at 25 -t 1°C and SO-S5% RH before iiiortality was assessed. Con- centration-mortality responses were estimated where sufficient females were available, with up to 16 concentrations and a water control. Control mortality was consistently <S%, even after 7 d. The criterjon for death was tlie inabilit>. to move more than its own body length when toiiclied with a fine paintbrush.

Crosses. After feniale p p a e from tlie 8 or S strain were reinovecl from rearing niediu, they were allowecl to emerge and inate witli inales from the opposite strain and to oviposit on diet (Brun et al. 1993). Resultant larvae were reared for mat- ing or testing. For the backcrosses, female FI pii- pie were removed from originul rearing media to new media and allowed to emerge ancl mate with males from either of tlie parental strains. Ovipo- sition occurred within a few days. The Fz generation was obtained in a siinilar fashion by allowvi~~g FI adults from a given cross to mate with one an- other.

Statistical Analysis. Abbott's (1925) forinIda was used to correct for natural mortality. Logit analysis was done with S-Plus (Statistical Sciences 1991). An F test was used to test for equality of the reciprocal FI responses. Responses in each of these crosses were overdispersed (larger than es- pected under the binomial model). This extra vari- ability was estimated by tlie heterogeneity (Statis- tical Sciences 1991) and used in analysis of deviance. The degree of dominance of resistance was calculuted from the formula of Stone (196S), with its standard errors calculated as described by Preisler et al. (1990).

Expected mortalities in the backcrosses were calculated assuming a single Iocus model, in which the mortality at any given dose was 0.5 x (FI mortality at that dose + backcross parent strain mortality at that dose) (Roush & Daly 1990, Tabashnjk 1991). The mortalities used for the FI and parent strains at any given concentration were those predicted by the appropriate regressions for each genotype. Although responses of the FI generation were significantly different, the differences were sinall and seemed unlikely to have biological sig- nificance. Therefore, the results were pooled to produce the ciirves of expected values in the Fig. 2. Expected values for the F2 were calculated sini- ilarly, as described later. To test whether tlie data fit a single Iocus model, a dispersion parameter was

* I' /. I , ' ' 1 ' 3 . 1 ' 4 , a ' 1 ' ,0,'000 ' a 100,000 ' 1 1 1 1 1

10 100 1,000 ENDOSULFAN CONCENTRATION (PPm)

Fig. 1. C;oiiceiitl.utiori-Inol.lalih. responses o f xhdt fe- inulc H. h t t t p i to eiitlosi~lf'an for sirsceptiblc a i d resistant ~iareiital strains aiid reciprocal FI hybrids after G 11 a i id i' tl. SI<. the offspriiig of siisceptilde ti.mdcs crossed with resistant inales. M, tlie offspring of resistiuit femnles crossed \vitIl sllsceptillle Inulcs.

first estimated by fitting a generalized additive niodel to tlie backcross responses. This method allowed 11s to model tlie resnonses iiori~arametricallv with a scatterplot smoothkg procedure (Chanibers fi- Hastie 1992).

Results and Discussion

Resistance Levels. During the initial selections, the niortality of feniales of the R parental strain after 'i d was 30-35% at 10,000 pprn of endosulfan and decreased slightly to 10% niortality during the coiirse of selection. In contrast, females of the S strain were all killed at 400 ppm even after only G li (Fig. 1, Table 1). Susceptible males appear to be even more sensitive than S females. Whereas the S feiilules liad an LCs0 of 25 ppni and 10% survived concentrations of 100 ppm at 6 li (Fig. l), among tlie S males, 73% tested ( 1 1 = 22) died at 35 ppm and none survived tests at 50 ppni. NO mortality was observed in males of the selected resistant strain, scored at 6 h after treatment at 400 ppni (Table 1).

Dominance of Resistance. When the FI feniales were tested, resistance from both reciprocal crosses was incompletely recessive to seniidomi- nant in inheritance, depending on the length of time before mortality was tallied (Fig. 1). When mortality was recorded after 6 li, the degree of dominance was -0.38 -+ 0.03 (S females X R inales) and -0.35 t 0.03 (R X S). After 7 d, D = -0.02 t 0.02 (S X R) arid D = -0.17 rt_ 0.02 (R

S), sug esting an increase in the rel. t' ive ex-

bioassay. The LCs0s of the two FI crosses were not significantly different based on overlap of the 95%

pression ( f ominance) of the resistance during the

. . i;

472 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. SS, no. 3

Table 1. Mortality of H. hanrpei by genotype and time at BOO ppm endosulfan

'i-d mortality f SEM n Genotype Genotype 11

fi-h mortality i SEM

S rriale 100 k 3.2 R ntale 3.Ï f 2.1 82 R male 0.0 k 0.9 u2 R(S)" m d e 3.3 k 3.3 30 S(R)'' male 100 c 1.4 49 SS female 100 f 0.4 lÏ0 SR" female 16.7 c 3.4 120

RR fernale 1.1 f 11.0 90 RS" female 82.2 * 4.0 Ï4

a Fis where the female parent is listed first. '>Tested at 50 ppm.

CL, but the hypothesis of equivalent responses was rejected for both 6 h and 'i d (P < 0.01).

Dominance expression is a phenotypic character liable to change with different types of bioassays (e.g., Roush & McKenzie 1987, Suckling et al. 1989, Roush & Daly 1990). Likewise, the length of time after treatment might also affect the expression of dominance. Nonetheless, resistance in this species shows the codominant inlieritance that is typical of resistance to cyclodiene insecticides (Brown 1967).

Sex Linkage. In contrast with females, we de- tected clear differences in responses of the FI males that depended on which parent was R, im- plying that resistance is sex-linked. At 400 ppm, a concentration of endosulfan that killed all S males and females at 6 11, all 49 sons of S females crossed with R males (S X R) died within 6 h (four replicates) (Table 1). Thus, R fathers did not influence the phenotype of their sons. In contrast, only 1 of the 30 sons of R females and S males died (five replicates), even after 7 d (Table 1). This result (3% mortality) is similar to the mortality observed for R males and females at the same concentration, but much less than for RS females (Table 1). The difference between males from reciprocal crosses was highly significant (P < 0.01, Wilcoxon rank test of the percentage mortality of the replicates).

Hypotlzenemus hainpei from Brazil were studied cytologically by Bergamin & Kerr (1953), who con- cluded that this species had an unusual sex-determination system. They reported that females had 12 autosomes (six pairs) and 2 X chromosomes, whereas males had 2 X chromosomes and 1 Y chromosome. Highly female-biased sex ratios, n = 10-12 females for every male (L.O.B. & V.G., unpublished data), were accounted for by loss of the Y chromosome through heteropycnosis in most of the sperm (i.e., the chromosome condenses into a densely staining mass, which is presumed to be inactive, and is eventually ejected from the cell). If Bergamin 87 Kerr's (1953) conclusions also apply to H. hampei populations from New Caledonia, our results suggest that the X chromosome con- tributed from the father is inactive in the sons, accounting for the phenotypic sex-linkage observed. However, recent observations suggest a more parsimonious hypothesis: a complete set of chromosomes may be inactive in the males (J.

Stewart & L.O.B., unpublished data). A similar example is known from the predatory mite, hfeta- se idus occidentalis (Nesbitt), where there is also a female-biased sex ratio (Roush & Hoy 1981) and where males are haploid because of tlie loss of the paternal set of chromosomes through heteropycnosis (Nelson-Rees et al. 19x0). For this species, loss of the paternal set was also demonstrated by insecticide resistance, which served as a genetic marker (Roush & Plapp 19S2). Thus, while the ex- act cytological system used by H. hairipei requires further resolution, our data strongly suggest that gene or chromosome inactivation is occuring.

Number of Loci Controlling Resistance. The data from all backcrosses were signficantly different (P e 0.01 even at 6 h, data not shown) from the expected values for a single gene (Figs. 2A and B). If resistance were due to a male haploid (or sex-linked) single major gene, the results for tlie FP should be different than if resistence were autosomal, because male haploidy (or sex-linkage) would tend to show a pattern similar to backcrosses to either the R or S parents. Autosomal inheritance (0.25 SS + 0.5 RS + 0.25 RR) would give a more intermediate response. The results for the FP appear to be more consistent with sex-linked (or male haploidy) inheritance rather than autosomal inheritance, since the F2 of S females by R males is much more suceptible than the F2 of the reciprocal cross. However, neither F2 shows a very close correspondence with the values expected (Fig. 2C).

Results from backcrossing experiments often give ambiguous results (Tabashnik 1991), and we suspect that these e.xperiments provide a further example of ambiguity. The interpretation of backcrosses is especially difficult where consistent results are difficult to achieve in bioassays.

A single major mechanism has been described that could account for all of the resistance observed. DNA sequencing has shown that a muta- tion in the Rdl GABA, receptor gene tlmt confers resistance to cyclodiene insecticides in Drosophila spp. (ffrench-Constant et al. 1993) and other in- sects also occurs in tlie Rdl homologue of H. hampei (ffrench-Constant et al. 1994). However, at present, we cannot discount the influence of other genes. The most direct way to test the influence of other genes would be by the use of genetic

I VOI. 88, no. 3 June 1995 B R U N ET AL.: ENDOSULFAN RESISTANCE I N H. hnlupei 473

predatory mite, hfetn- ), where there is also a d i & Hoy 1981) and

nancial assistance for D.M.S., the Dean of the College of Agriculture at Cornell University for travel support for

(1953), and I'. Borsa, P. Gingericli, and J. Stuart for useful comments.

I I 90 R.T.R., P. Gingericli for translating Bergainin K- Kerr brtality I;EM

' 2.1 82

30 lj 3.3 30

also demonstrated by li served as a genetic 2 ) . Thus, while the ex- by H. hanipei requires . strongly suggest that ntion is occuring. lling Resistance. The

if resistence were au- loidy (or sex-linkage) i similar to backcross- nts. Autosomal inher- 0.25 RR) would give

99.1

90.

70- 50: 30.

the values expected

/ 1 , , , , . / I , , , , , , , ( , , , , , . , ) , , , , , , , , , I 10 100 1,wo l0,bW 1"

EN DOS U LFA N CONCENTRATI ON ( P P m )

Fig. 2. Coticentrution-inortalih responses of d i d t female H. / i m t i p i to erit~osiilfm after Ï d for hcltcrosses of FI fein&s to ( A ) resistant indes m d (ß) snsceptilik males. Fiilely tlaslied C I I I T ~ S sliow the expected responses if resistance resulted from a single gene. Allele intlicated in parentlieses, e.g., (13). intlicntes the allele that is ap- parently inactivated or lost in the niules, its tliscussetl in the text. Resiilts for F., adiilt females are S I I O M ~ in C, wliere the midtlle, dotted cmye sliows the expected values asstinting that the males were Iieteroqgotes (i.e., tlie pntemal allele coiiltl be passed v i t a the FI moles to their Fe offspring) and the outer dashed cun'es show the ex- pectatioiis if males were lieiniygotes (i.e., tlie FI males transmitted only the muternal dlelei.

markers so that tlie infliience of all chromosomal regions (as done by McKenzie et al. 1980) could be assessed.

Two salient points emerge from these studies; first, the high levels of resistailce found in the FIS (Fig. l), which are heterozygous for at least one major gene ( E d ) , suggest that heterozygotes could survive field exposure ( R O L I S ~ & Dalv 1990). This would help to account for the rapid changes in resistance frequency which were reported by Brun & Suckling (1993). Field bioassays (e. Parkin et

to investigate this further. Second, the strong a - arent sex-linkage of resistance sug ests, for tl!e

chroniosome or (more likely) the entire set of paternal chromosomes is inactivated in males.

I

1

I

al. 1999) of known genotypes wouldfk required

Erst tirne in this species, that either t f le paternal X

Acknowledgments \Ve tliank the Coffee Board (ADRAF) for their su port

of this project, the French Embassy (LVellington) g r fi-

References Cited

Ahliott, W. S. 1925. A method for computing tlie ef- fectiveness of an insecticide. J. Econ. Entomol. 18: 265-26;.

Bergainin, J. & W. E. Kerr. 1953. Determincao do sexo e citologia d:i broca do cafe. Ciencia e Cultura. 3: 11;-191.

Brown, A.W.A. 1967. Genetics of insecticide resistance in insect vectors, pp. 505-557. Iti J. U: IL'right & R. l'al [eds.], Genetics of insect vectors of disease. Elsevier, New York.

Brun, L. O. 8; D. M. Suckling. 1992. Field selection for entlosulfan resistance in H!/potlìerrett~rrs Iìattipei (Coleopterx Scol!tidae) in New Cnledonia. J. Econ. Entomol. 85: 32*5-:334.

Brun, L. O., C. Marc.illauc1, V. Gaudiehon & D. M. Suc-kling. 1989. Entlosnlfh resistance in H!/po- tlietiet~iris l i u ~ t ~ p i (Coleoptera: Scolytidae) in Nekv Caletloniu. J. Econ. Entomol. 8-7: 1311-1316.

1990. Monitoriiig of endosiilfan and lindane resistance i i i tlie coffee berq. borer, Hyl"~tl2etlettrrls l i o t t ì p i (Co- leoptera: Scolytitlae) in Ne\\. Culedonia. Hull. Ento-

199 1. E\zdriation of a rapid bioussuty for endosulfan resistance in coffee beriy borer, H y p o t l i e t ~ e n ~ z ~ ~ hatti- pi (Ferrari) (Coleoptera: Scolytidae). Trop. Pest hlanage. 3;: 221-223.

Br~ni, L. O., V. Gaudichon & P. J. Wigley. 1993. An artificial diet for continuous rearing of coffee berry borer, H ~ ~ ~ ( i f / i e n e ~ t t r r s hanipei (Fei-rari) (Coleoptera: Scolytitlue). Insect Sci. Appl. 14: 5S5-558.

ffrench-Constant, R. H., J. C. Steichen & L. O . Brun. 1994. A molecular diagnostic for endosulfan resistance in tlie coffee berry borer H!p~tAenenzu.s hnmpei (Coleoptera: Scolytidae). Bull. Entomol. Res. S4: 11- 15.

ffrench-Constant, R. H., J. Steichen, T. A. Roche- leau, K. Aronstein & R. T. Roush. 1993. A single- amino acid substitution in it gamina-aininobutc acid subtype A receptor locus is associated with cyclodiene insecticide resisttnce in Dt-o.r-opllila populations. Proc. Nat. Acitd. Sci. 90: 1957-1961.

IMcICenzie, J. A., J. RI. Dearn & $1. J. Wiitten. 1980. Genetic basis of resistance to diazinon in Victorian populations of the Australian sheep blowfly, Lucilin crrprinn. Aust. J. Biol. Sci. 33: S5-95.

Nelson-Rees, W. A., M. A.-Hoy & R. T. Roush. 1980. Heterocllroinatinization: chromatin elimination and haploidizition in the parahaploid mite hfetuseiulus oc- cideiitcili.~ (Nesbitt) (Acarina: Phytoseiidae). Chromo-

Parkin, C. S., L. O. Brun & D. RX. Suckling. 1992. Spray deposition in relation to endosulfan resistance in the coffee berry borer Hyptliene~nus hnmpei (Co- leptera: Scolytidae) in New Caledonia. Crop Prot. 11: 913-220.

Potter, C. 1952. An improved laboratory apparatus for ;tpplying direct sprays and surface films, with data on the electrostatic charge on atomized spray fluids. Ann. Appl. Biol. 38: 1-2s.

I l K J l . HeS. 80: 1299-135.

~011iil ÏÏ: 263-27G.

4ï4 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 88, no. 3

Preisler, H. K., RI. A. Hoy & J. L. Robertson. 1990. Statisical analyses of niodes of inheritance for pesticide resistance. J. Econ. Entomol. 83: 1649-1665.

Roush, R. T. & J. C. Daly. 1990. The role of popu- lation genetics in resistance research and manuge- ment, pp. 97-152. I n R. T. Rous11 & B. E. Tabuslinik [eds. J. Pesticide resistance in artliropocls. Chapman and Hall, New York.

Roush, R. T. & J. A. McKenzie. 1987. Ecological genetics of insecticide and acaricide resistance. Annu. Rev. Entomol. 32: 361350.

Roush, R. T. & F. W. Plapp, Jr. 1982. Biochemicnl genetics of resistance to aryl carbainuite insecticides in the predaceous mite, Alefasc~iri~ris occidentdis. J. Econ. Entomol. 7!5: 304307.

Statistical Sciences. 1991. Reference manual, version 2.0. SbatSci, Seattle, \Vash.

Stone, B. E 1968. A formula for determining the degree of dominance in cases of monofactorial inheritance of resistance to chemicals. Bull. IV.H.0. 38: 325326.

Suckling, D. RI., J. Khoo & D. J. Rogers. 1989. The dynamics of azinpliosmetliyl resistance in E p i p h y m postnittclna. J. Econ. Entomol. 82: 1003-1010.

Tahashnik, B. E. 1991. Determining the mode of inheritance of pesticide resistance from backcross experiments. J. Econ. Entomol. S 4 703-712.

Receiccclfor pchliccition 28 Februciry 1994; accepted 1 Dcceniber 1994.

I

US ISSN 0022 0493

J U N E 1995 NUXIBER :3

Qp1 z?f ilished by THE ENTOMOLOGICAL SOCIETY OF AMERICA

genetics of endosulfan resistance in hypothenemus hampei...

Documents