Genetics and linkage of aldrin resistance in the German cockroach, Blattella germanica (L.)*

Bull. Org. mond. Sante 1969, 40, 745-752 Bull. Wld Hlth Org. Genetics-and Linkage of Aldrin Resistance in the German Cockroach, Blattella germanica (L.) * I. C. McDONALD,1 M. H. ROSS 2 & D. G. COCHRAN 3 The inheritance mechanism and linkage association of cyclodiene resistance were studied in 2 cyclodiene-resistant strains of the German cockroach. The resistant strains were crossed to each other, to an unmarked susceptible strain, and to 6 susceptible marker strains. The progeny from succeeding generations were testedfor resistance to aldrin using a dipping method or a time-exposure method, and were examined for markers in linkage tests. The results show that cyclodiene resistance (R-Cyclo) is inherited as a simple auto- somal semidominant trait which is allelic in the 2 resistant strains. Linkage studies indicate that R-Cyclo is independent of linkage groups Il-VI, but is linked with the group VII trait curly wing at a distance of 3.4±1.0-6.2±1.4. The physiological basis for cyclodiene resistance in insects has not yet been satisfactorily explained. However, genetic studies have shown that this type of resistance is inherited in patterns consistent with Mendelian principles. A single semidominant auto- somal factor for dieldrin resistance has been demon- strated for many insect species (Crow, 1957; David- son, 1965; Davidson & Mason, 1963; Georghiou, 1965; Klassen, 1966; Milani, 1960, 1963; Oppe- noorth, 1965). Dominant cyclodiene resistance has also been reported (Davidson, 1965; Davidson & Hamon, 1962; Rozeboom & Johnson, 1961; Stone, 1962). Other studies have indicated that some insects have a multifactorial inheritance pattern (Abdullah, 1961; Abedi, 1958; Barbesgaard & Keiding, 1955; Bragassa & Brazzel, 1961; Busvine & Khan 1955; Grayson, Jarvis & Levitan, 1956; Jarvis, Grayson & Levitan, 1957; Rahman & Khan, 1964), but some questions remain concerning the interpre- tation of the results. Among the latter group is the German cockroach, Blattella germanica (L.), upon which additional genetic studies have been carried opt in this laboratory. The purpose of this paper is * From the Department of Entomology, Virginia Poly- technic Institute, Blacksburg, Va., USA. This investigation was partially supported by Grant No. CC-00264-10 from the National Communicable Disease Center, US Department of Health, Education and Welfare, Atlanta, Ga., USA. 1 Present address: Metabolism and Radiation Research Laboratory, Agricultural Research Service, US Department of Agriculture, Fargo, N. Dak., USA. 'Instructor in Entomology, Virginia Polytechnic Institute. Professor of Entomology, Virginia Polytechnic Ins- titute. to present data on the mode of inheritance of aldrin resistance, and to report the results of linkage tests between aldrin-resistant stocks and selected mutant markers in this species. A preliminary note on some of this work has already appeared (Cochran, 1965). MATERIALS AND METHODS The German cockroaches employed in these experiments were from cultures of susceptible, cyclo- diene-resistant, and mutant stocks reared continu- ously in this laboratory. The susceptible strain has been maintained for many years without intentional exposure to insecticides. The resistant strains (Aldrin R and Landstuhl R) were selected with aldrin on a periodic basis to ensure stock integrity and to maintain a high level of resistance. The mutant stocks used for linkage tests were balloon wing (ba)-group II; rose eye (ro)-group III; orange body (or)-group IV; glassy wing (gl)- group V; black body (Bl)-group VI; and curly wing (Cu)-group VII. No sex-linked trait (group I) was included since Cochran (1965) reported that aldrin resistance in this species is not sex-linked. The visible mutants have been described elsewhere (Cochran & Ross, 1961, 1967; Ross & Cochran, 1962, 1966), but briefly ba, ro, or and gl are inherited as simple auto- somal recessives. B! is a semidominant autosomal trait with a distinct heterozygote. Cu is inherited as a semidominant autosomal factor which is lethal in the homozygous condition. Cu females are sterile and decreased viability is apparent in outcrosses involv- ing Cu males. The cultures were reared by standard 2331 - 745 - 7. C. MCDONALD, M. H. ROSS & D. G. COCHRAN methods developed in this laboratory (Clarke & Cochran, 1959; Grayson, 1951), except that 1 US quart or 1 US gallon (approx. 1 litre and 3.75 litres) glass containers were used exclusively to house the insects. Crosses were made reciprocally between aldrin- resistant and wild-type (susceptible) or mutant cockroaches except those involving Cu, where female sterility precluded reciprocal crossing. To establish the inheritance mechanism for aldrin resistance F1, F2 and backcrosses to resistant and susceptible parental stocks were made. In addition, the 2 resistant strains were crossed reciprocally for 3 generations to test for allelism. For linkage studies, the same general approach was used except that backcrosses were usually made to the recessive mutant stock. Where Cu was involved, F1 males exhibiting the trait were backcrossed to the suscep- tible strain. A scarcity of F1 Cu males, resulting from reduced viability, prevented backcrosses with the resistant parent, while Cu female sterility made F2 crossing impossible. With Bl, F1 and F2 crosses as well as crosses to the normal susceptible strain were made. In most instances mass matings were carried out because the basic inheritance pattern of the mutants was known and no particular problems were anticipated. With Cu, however, this was not so, and some individual crosses were made. The in- tegrity of the matings was ensured by separating the sexes in the nymphal stage. Data from the crosses were subjected to the x2 goodness-of-fit test to ana- lyse variations from expected ratios. Resistance testing was conducted using a dipping technique or a continuous time-exposure method. Technical-grade aldrin (95%) supplied by the Shell Chemical Company was used for testing. Regard- less of the method, the insecticide was first dissolved in redistilled acetone to form stock solutions of 1 % or 5 %. With the dipping technique, suspensions of appropriate concentrations were prepared and used as previously described (Clarke & Cochran, 1959). For the time-exposure method, 3 ml of stock solution were applied evenly to a 200-cm2 piece of Whatman No. 1 filter-paper several hours prior to use. The 1% or 5% stock solutions were used for male and female tests, respectively, since adult females withstand more aldrin than do adult males. The insects were confined to the treated surfaces in groups of 35-45 and mortality counts were made periodically for 80 hours. The dipping technique was used to establish baselines for the susceptible and resistant cock- roaches and for the progeny of crosses between them. Last-instar nymphs were trea ed to obtain dosage- mortality data which were fitted to regression lines by the method of Bliss (1935). Each line was based upon at least 5 points which had been replicated 3-7 times. Each replicate consisted of approximately 15 insects. Discriminating concentrations were determined from the lines and were used in subsequent tests to separate genotypes. The dipping technique was also used for linkage studies with the eye colour and body colour mutants. However, the method was inade- quate for certain of the wing mutants which appear only in adults. The time-exposure method proved satisfactory for such adult traits. Indeed, discri- minating times could be established for both males and females using the time-exposure method. Because of this fact, the results of dipping tests were confirmed by time-exposure tests whenever cockroaches were still available. Adults were tested 7-14 days following metamorphosis. Mutant stocks were also exposed to aldrin to test for susceptibility. In every case the response was similar to that of the susceptible strain. RESULTS Strain response to aldrin Fig. 1 and 2 show the bases for separation of genotypes by the 2 test methods employed in this work. The best separation occurred between the susceptible and heterozygous types with both test methods. A discriminating concentration of 0.07 g/litre for nymphs was determined from Fig. 1. Corresponding data for adult females produced a value of 0.04 g/litre. Overlap in the lines precluded the use of a discriminating concentration to separate heterozygous from resistant individuals. Likewise, all the lines from data with adult males overlapped to the extent that the results were unsatisfactory. By the time-exposure method all 3 male genotypes can be separated successfully (Fig. 2). Comparable data for adult females agree essentially with those for males, although some overlap occurs at each of the 2 juncture points. Thus, the best results were achieved here too by selecting an exposure time for separating susceptible from heterozygous types. The times are 8 hours for adult males and 14 hours for adult females. Inheritance mechanism Table 1 shows the results of dipping tests with aldrin on parental types and F1 progeny from 746 INHERITANCE OF ALDRIN RESISTANCE IN THE GERMAN COCKROACH FIG. 1 MORTALITY IN 3 STRAINS OF GERMAN COCKROACH NYMPHS FOLLOWING DIPPING TREATMENT WITH SUSCEPTIBLE / 95 STRAIN 90n ,/ F,HETEROZYGOTE ,_o 80 - 70 I; 60 50 p= 40 30 20 10 5 r - II I II I I I I Iutu ALDRIN 1.0 CONCENTRATION (g/I) MORTALITY IN 3 STRAINS OF FIG. 2 ADULT MALE GERMAN COCKROACHES FOLLOWING TO FILTER-PAPER TREATED WITH 1 % ALDRIN CONTINUOUS EXPOSURE ALDRIN - RESISTANT STRAIN 202 tested -L s3 5 7 1 1 9 15 21 TIME (hours) I1 |S24 44 64 > om3 90f669 0.01 01 10.0 wH0 90668 100 r FlHETERI 114te,sted F-_ 90 80 70 so I_- 50 =40 30 20 10 L4' 8 ITIBLE AIN :,sted 747 SUSCEF STR 182 ti 1. C. MCDONALD, M. H. ROSS & D. G. COCHRAN TABLE I RESPONSE OF PARENTAL AND F, STRAINS OF GERMAN COCKROACHES TO ALDRIN IN DIPPING TESTS ON LARGE NYMPHS reciprocal crosses between the susceptible strain and the aldrin-resistant strain. It is clear from this table, and the results presented in Fig. 1 and 2, that the F1 hybrids are intermediate in resistance between the parental types. The close similarity of the 2 groups of F1 hybrids in LC50, slope, and times-resistance indicates autosomal inheritance. The F2 and back- crosses were also assessed by the discriminating- concentration technique (Table 2). The findings are those expected on the basis of a 1-factor autosomal Mendelian trait. The intermediate F1 hybrids were not tested by the discriminating concentration of 0.07 g/litre, but the dosage-mortality data indicate they would have survived this treatment (Fig. 1). Corresponding crosses were also made between the susceptible strain and the Landstuhl resistant strain. This latter strain has an LC50 of 4.3 g/litre and a times-resistance factor of 310 at LC50. The crossing TABLE 2 RESULTS OF TESTING NYMPHAL PROGENY FROM SEVERAL CROSSES FOR SUSCEPTIBILITY TO ALDRIN BY USE OF A DISCRIMINATING CONCENTRATION a Percentage of population susceptible Strain to aldrin No. tested Expected Observed S? x RS F2 25 24.7 332 Ri x Sd F2 25 25.9 135 (R?ixSci?)FaxR 0 0 164 (S?x Rc) F. x R 0 0 169 (R?x Sd) FI x S 50 47.8 301 (S?ix Rd) F,x S 50 49.5 271 a 0.07 g/l. results were essentially identical with those reported for the Aldrin strain, even though the Landstuhl strain has a lower resistance level. The F1 hybrids were slightly less resistant (LC5e=0.45 g/litre) and produced a times-resistance factor of 32 for both reciprocal crosses. The F2 and backcross to sus- ceptible showed approximately 25% and 50% susceptibility to the discriminating dose of aldrin, respectively. Thus, it appears that a 1-factor autosomal inheritance mechanism is also operating in the Landstuhl strain. Tests for allelism were undertaken by crossing the 2 resistant strains. Table 3 gives the results of tests TABLE 3 RESULTS OF CROSSING EXPERIMENTS BETWEEN THE ALDRIN- AND LANDSTUHL-RESISTANT STRAINS IN DIPPING TESTS ON LARGE NYMPHS Strain LCso g/l Times-resistance Aldrin-R 7.2 514 Landstuhl-R 4.3 310 (A? x LS) F, 8.2 585 (LY x Ad) Ft 5.8 415 (AY x Ld) F2 7.2 514 (L? x Ad) F2 6.8 493 (A9 x Ld) F3 5.6 400 (Li? x Ad) F3 5.4 390 with the parental stocks and their resulting F1, F2 and F3 progenies. No diminution in resistance level occurred in any of the crosses. If the factors were non-allelic, heterozygous and susceptible, individuals would be expected to appear in the F2 and subse- quent generations. A discriminating concentration of 1.0 g/litre produced no mortality among the F2 or F3 progeny from either cross. Indeed, the most prominent features of these results are the apparent heterotic effect in the F1 progeny of A9 x LS and the convergence of the resistance level in both sets of the F2 and F3 progeny. The drop in Fs LC50 values may be attributable to sampling or testing variability, but does not indicate a loss of the homozygous' genotype. While the possibility of closely linked additive factors has not been completely eliminated, the most logical interpretation of the results is that the resistance factors are allelic in the 2 strains. 748 INHERITANCE Of ALDRIN RESISTANCE IN THE GERMAN COCKROACH Linkage studies Tests for linkage were made by crossing the aldrin- resistant strain with susceptible mutant stocks. These tests involved markers on linkage groups II- VII as indicated previously. The results are pre- sented in Table 4. Earlier linkage studies with the group II trait ba gave inconclusive results (Cochran, 1965), primarily because the dipping technique was inadequate in separating genotypes. The present data (Table 4, crosses 1 and 2) were obtained with the continuous- exposure method and they show independent assortment of R-Ald and ba. The ba factor is believed to be linked with r-DDT(Cochran, 1965; Cochran & Ross, 1962). Therefore, it would appear that the factors for R-Ald and r-DDT are located on different chromosomes. The group III factor ro also appears to segregate independently of R-Ald even though a significant deficiency of the ro phenotype occurred in both the test-cross and F2 progenies (Table 4, crosses 3 and 4). The penetrance of ro has been good in previous studies (Ross & Cochran, 1966; 1967) and the present deficiency of ro individuals cannot be explained. Nevertheless, the deficiency is not what would be expected in a case of linkage. Accordingly, a con- tingency table (Mather, 1957) was employed to make allowance for the disturbed single factor ratio. The subsequent X2 values obtained for the data of crosses 3 and 4 (Table 4) clearly imply independent assortment of ro from R-Ald. Cochran (1965) reported independent assortment of R-Ald from the group III traits red eye (r) and pro-wing (Pw). Pw is associated with a reciprocal chromosome transloca- tion (Cochran & Ross 1969), and also marks linkage group VIII (Ross & Cochran 1968b). Thus, R-Ald appears to be independent of linkage group VIII as well. Results with the group IV trait or (Table 4, crosses 5 and 6) and the group V trait gl (Table 4, crosses 7 and 8) indicate that these factors segregate TABLE 4 DATA FROM LINKAGE TESTS BETWEEN ALDRIN-RESISTANT AND MUTANT STOCKS OF BLATTELLA GERMANICA (L.) Matings 1. bajba+,R-Ald/R-Ald+ xba/ba,R-Ald+/R-A1d+ 2. ba/ba+,R-Ald/R-Ald+ x ba/ba+,R-Ald/R-Ald+ 3. ro/ro+,R-Ald/R-Ald+ x ro/ro,R-Ald+/R-Ald+ 4. ro/ro+,R-Ald/R-Ald+ x ro/ro+,R-Ald/R-Ald+ 5. or/or+,R-Ald/R-Ald+ x or/or,R-Ald+/R-Ald+ 6. or/or+,R-Ald/R-Aid+ x or/or+,R-Ald/R-Ald+ 7. g1/g1+,R-Ald/R-Ald+ x gllg1,R-A1d+1R-A1d+ 8. g/IDI+,R-Ald/R-Ald+ xgl/gl+,R-Ald/R-Ald+ 9. Bl/B1+,R-Ald/R-Ald+ x Bl+/Bl+,R-Ald+/R-Ald+ 10. B1/Bl+,R-Ald/R-Ald+- x B1/B1+,R-Ald/R-Ald+ 11. Cu R-A/d+/Cu+ R-Aldl x Cu+ R-Ald+/Cu+ R-Ald+ Phenotypes and numbers of progeny 235 ba+,R-Ald 811 ba+,R-Ald 467 ro+,R-Ald 1 247 ro+,R-Ald 271 or+,R-AId 662 or+,R-Ald 580 gl+,R-Ald 1 366 gI+,R-AId 444 Bl+,R-Ald 639 Bl C,R-Ald 151 Cu+,R-Ald 198 ba+,R-Ald+ 246 ba+,R-Ald+ 461 ro+,R-Ald+ 448 ro+,R-Ald+ 317 or+,R-Ald+ 240 or+,R-Ald+ 589 gi+,R-Ald+ 477 g/+,R-Ald+ 410 Bl+,R-Ald+ 215 Bl+,R-Ald 0 Cu+,R-Ald+ 225 ba,R-Ald 272 ba,R-Ald 349 ro,R-Ald 322 ro,R-Ald 311 or,R-Ald 251 or,R-Ald 541 glR-Ald 449 gIR-Ald 385 Bl b,R-Ald 254 B C,R-Ald+ 10 Cu,R-Ald X2 4.33 3.45 0.3X 10-sa 0.39 a 4.93 5.85 4.22 1.64 4.30 5.94 210.93 203 ba,R-Ald+ 101 ba,R-Ald+ 342 ro,R-Ald+ 107 ro,R-Ald+ 318 or,R-Ald+ 91 or,R-Ald+ 532 gl,R-Ald+ 163 glR-AId+ 419 BI b,R-Ald+ 72 Bi+,R-Ald+ 135 Cu,R-Aid+ p >0.20 >0.30 >0.99 >0.80 >0.10 >0.30 >0.60 >0.20 >0.10 <0.001 ci a Single degree of freedom; x I based upon contingency test. b Heterozygous for B. c Bi,R-Ald and BI,R-Ald+ classes contain both heterozygous and homozygous individuals for Bl. d Significant deviation from ratio expected for independent assortment (1: 1:1:1). 749 1. C. MCDONALD, M. H. ROSS & D. G. COCHRAN independently from R-Ald. The relatively low num- ber of or+, R-Ald individuals shown in cross 5 does not signify linkage as this class represents a parental combination of factors. The results confirm a previous report of the independence of or and R-Ald (Cochran, 1965). The poor fit of the or data to expected ratios is probably attributable to the use of the dipping technique with which data were collected on females only. The data on gl were obtained by the continuous exposure method and they fit the expected ratios more closely. Test results with the group VI trait BI are pre- sented as crosses 9 and 10 in Table 4. Both sets of data show the independent assortment of BI and R-Ald. The relatively poor fit of the data to expected ratios can again be attributed to the use of the dip- ping technique. Also, body colour was found to deteriorate in moribund cockroaches. Precautions were taken to avoid this difficulty but a small pro- portion of misclassified specimens may be present in the initial data which are included in the tabulations. Only backcross data were obtained with the group VII trait, Cu. The results of these tests are shown in Table 4, cross 11, and clearly indicate linkage between Cu and R-Ald. Also shown is the complete absence of the normal-susceptible cross- over class. This absence could be fortuitous or it may be related to the involvement of Cu with a reciprocal chromosome translocation (Cochran & Ross, 1969). Suppression of the crossover class might occur as a result of some complication occur- ring during meiosis in the Cu translocation hetero- zygote. Possibilities of this sort have a bearing on the calculation of linkage distance. If suppression of one crossover class is assumed, then allowance for this class should be made. The calculated linkage distance is 6.2±1.4 map units when this is done. Contrarily, if the absence of one class of individuals is assumed to be fortuitous, no allowance need be made and the linkage distance is 3.4±1.0 map units. Since additional data will be required to distinguish between these possibilities, the range of values should be used at present in designating the linkage distance between Cu and R-Ald (6.2±1.4-3.4±1.0). With each of the mutant stocks used in this work, backcrosses were made between the heterozygous F1 and the aldrin-resistant strain. The progeny were challenged with a discriminating treatment of aldrin to determine their resistance. All of the progeny would be expected to survive the challenge since they should be either heterozygous or homozygous for the resistance factor. Therefore, these crosses served as a further check on the purity of the aldrin- resistant strain. Approximately 2800 progeny were tested in this manner with a resultant mortality of less than 1%. This small percentage of mortality probably does not give a true indication of strain impurity because insect manipulation in control- type experiments often results in 1 %/7-3% mortality. These results, which are based upon the original involvement of 150-200 individuals from the aldrin- resistant stock, are a further indication that the resistance factor is present in this stock at a high gene frequency, although probably not at fixation. DISCUSSION The results show that cyclodiene resistance (as challenged by aldrin) in the 2 German cockroach strains is inherited as a simple autosomal semidomi- nant trait. This finding does not agree with the interpretation of corresponding results presented by Grayson, Jarvis & Levitan (1956) and Jarvis, Grayson & Levitan (1957) using chlordane as the challenging agent. It should be pointed out that their work was reported prior to the widespread use of the discriminating-challenge technique, and was based largely upon dosage-mortality response curves. However, a review of their published data, particu- larly those of Grayson, Jarvis & Levitan (1956), reveals a pattern which is in basic agreement with the results presented here. The work of Jarvis, Grayson & Levitan (1957) on F2 crosses is not pre- sented in sufficient detail to make an interpretation in terms of present-day concepts. Nevertheless, it is our opinion that there is no longer any basis for maintaining that the major locus for cyclodiene resistance in these strains of the German cockroach is inherited by any mechanism other than a simple autosomal semidominant factor. Accordingly, it is proposed that this factor be referred to as R-Cyclo in the future. The finding of linkage between Cu and R-Cyclo has several important connotations. As already mentioned, it shows that the 2 types of resistance to chlorinated hydrocarbon insecticides are apparently located on different chromosomes. This is similar to the findings on houseflies (Plapp & Hoyer, 1967) and Culex pipiens fatigans Wied. (Tadano & Brown, 1967) but is contrary to those with Anopheles quadrimaculatus Say (Davidson, 1965) and Aedes aegypti (L.) (Klassen & Brown, 1964) where the factors are linked. In the German cockroach, the linkage situation is 750 INHERITANCE OF ALDRIN RESISTANCE IN THE GERMAN COCKROACH 751 also of interest because the factor with which R-Cyclo is linked is involved with a reciprocal chromosome translocation. Indeed, the 2 chromo- somes taking part in the Cu translocation have been identified as the shortest and the second longest autosomes (Cochran & Ross, 1969). Therefore, it is expected that Cu will be associated with 2 linkage groups. No trait other than R-Cyclo has yet shown linkage with Cu. Because Cu has already been pro- posed as a marker for group VII (Ross & Cochran, 1968a), we are tentatively assigning R-Cyclo to group VII as well. Additional traits linked with Cu but not with R-Cyclo will be placed in a different linkage group. The linkage results with Cu and R-Cyclo are of interest in another connexion. They represent an additional example of close linkage. Linkage studies reported to date for this species show a preponder- ance of close linkage. It is not yet possible to determine whether this is fortuitous or has meaning with regard to the arrangement of euchromatin along the chromosomes of this insect. A comment is in order on the comparison of the 2 test methods used in this study. It is clear that the time-exposure method is to be preferred when genetic analyses are involved. In general, the results gave a better fit to the expected ratios by this method than by the dipping technique. Indeed, with further prac- tice, it is probable that all 3 genotypes could be separated routinely by the time-exposure method. RESUMt GINETIQUE ET TRANSMISSION SOLIDAIRE DE LA RtSISTANCE A L'ALDRINE CHEZ LA BLATTE, BLATTELLA GERMANICA (L.) Utilisant un certain nombre de souches de Blattella germanica, dont des souches marqueuses et deux souches tr&s resistantes, les auteurs ont analyse le mecanisme de la transmission et les modalites de la liaison gen6tique de la resistance a l'un des derives du cyclodiene, l'aldrine, chez cet insecte. On a a cet effet procede a des croisements entre les souches resistantes, une souche sauvage sensible et six souches marqueuses sensibles (groupes II a VII), puis mesure la resistance a l'aldrine dans la descendance grace a une technique d'immersion ou par exposition continue a l'insecticide. La premiere technique a permis d'etablir des lignes de r6gression dose-mortalite pour chacun des differents lots d'insectes et de definir les doses discriminant les individus sensibles et les individus r6sis- tants homozygotes et heterozygotes. L'emploi de la seconde methode a conduit de meme a fixer des durees d'exposition discriminantes. Les resultats montrent que la resistance aux derives du cyclodiene (R-cyclo) est transmise comme un caractere simple autosomique et semi-dominant dans les deux souches resistantes. Ils corroborent les observations faites chez un certain nombre d'especes d'insectes. Des croise- ments entre les souches resistantes ont mis en evidence l'all6lisme des genes. Quant aux etudes de linkage, elles ont e men6es sur les groupes II a VII porteurs de genes marqueurs; le gene de la r6sistance etant par nature autosomique, on n'a pas utilise le groupe I porteur d'un marqueur lie au sexe. I1 ressort de ces etudes que la resistance aux derives du cyclodiene n'est pas liMe aux genes marqueurs caracterisant les groupes II a VI, mais est liee en revanche au gene marqueur du groupe VII, Cu (curly wing). Un des types d'enjambement qui aurait du resulter des croisements n'a pas ete observe. Cette lacune, ainsi que le fait que le gene Cu est impli- que dans un phenomene de translocation chromo- somique reciproque, n'autorise qu'une evaluation tres approximative de la position respective des genes en cause. REFERENCES Abdullah, M. (1961) J. Hered., 52, 179-182 Abedi, Z. H. (1958) Bull. ent. Res., 49, 637-642 Barbesgaard, P. & Keiding, J. (1955) Vidensk. Meddr. dansk naturh. Foren., 117, 83-116 Bliss, C. I. (1935) Ann. appl. Biol., 22, 134-167 Bragassa, C. B. & Brazzel, J. R. (1961) J. econ. Ent., 54, 311-314 Busvine, J. R. & Khan, N. H. (1955) Trans. roy. Soc. trop. Med. Hyg., 49, 455-459 Clarke, T. H. & Cochran, D. G. (1959) Bull. Wid Hlth Org., 20, 823-833 Cochran, D. G. (1965) Insecticide resistance factors and genetic linkage in Blattella germanica (Dictyoptera). In: Proceedings of the 12th International Congress of Entomology, London, 1964, p. 240. Cochran, D. G. &Ross, M. H. (1961) Va J. Sci., 12, 10-17 Cochran, D. G. & Ross, M. H. (1962) Bull. Wld Hlth Org., 27, 257-261 752 1. C. MCDONALD, M. H. ROSS & D. G. COCHRAN Cochran, D. G. & Ross, M. H. (1967) Cockroach gene- tics. In: Wright, J. & Pal, R., ed., Genetics of insect vectors of disease, Amsterdam, Elsevier, pp. 403-415 Cochran, D. G. & Ross, M. H. (1969) J. Hered. (in press) Crow, J. F. (1957) Ann. Rev. Ent., 2, 227-246 Davidson, G. (1965) Resistance to chlorinated insecticides in anopheline mosquitos. In: Proceedings of the 12th International Congress of Entomology, London, 1964, pp. 236-237 Davidson, G. & Hamon, J. (1962) Nature (Lond.), 196, 1012 Davidson, G. & Mason, G. F. (1963) Ann. Rev. Ent., 8, 177-196 Georghiou, G. P. (1965) Adv. Pest Control Res., 6, 171-230 Grayson, J. M. (1951) J. econ. Ent., 44, 315-317 Grayson, J. M., Jarvis, F. E. & Levitan, M. (1956) J. econ. Ent., 49, 130-131 Jarvis, F. E., Grayson, J. M. & Levitan, M. (1957) J. econ. Ent., 50, 185-187 Klassen, W. (1966) Mosquito News, 26, 309-318 Klassen, W. & Brown, A. W. A. (1964) Canad. J. Genet. Cytol., 6, 61-73 Mather, K. (1957) The measurement oflinkage in heredity, New York, John Wiley Milani, R. (1960) Misc. Pubi. ent. Soc. Amer., 2, 75-83 Milani, R. (1963) Bull. Wld Hith Org., 29, Suppl., pp. 77-87 Oppenoorth, F. J. (1965) Ann. Rev. Ent., 10, 185-206 Plapp, F. W. & Hoyer, R. H. (1967) J. econ. Ent., 60, 768-774 Rahman, S. J. & Khan, N. H. (1964) Botyu-Kagaku, 29, 19-21 Ross, M. H. & Cochran, D. G. (1962) Nature (Lond.), 195, 518-519 Ross, M. H. & Cochran, D. G. (1966) J. Hered., 57, 221-226 Ross, M. H. & Cochran, D. G. (1967) J. Hered., 58, 274-278 Ross, M. H. & Cochran, D. G. (1968a) J. Hered., 59, 105-110 Ross, M. H. & Cochran, D. G. (1968b) J. Hered., 59, 318-320 Rozeboom, I. E. & Johnson, R. (1961) Amer. J. trop. Med. Hyg., 10, 775-781 Stone, B. F. (1962) Aust. J. agric. Res., 13, 1008-1022 Tadano, T. & Brown, A. W. A. (1967) Bull. Wld Hlth Org., 36, 101-111