Heredity 75(1995) 4 16—424 Received22 March 1995 Genetic differentiation in Zeiraphera diniana (Lepidoptera: Tortricidae, the larch budmoth): polymorphism, host races or sibling species? IGOR EMELIANOV*, JAMES MALLET & WERNER BALTENSWEILERt Ga/ton Laboratory, Department of Genetics and Biometry, University College London, 4 Stephenson Way, London NWI2HE, U.Kand Blumenbergstrasse 9, CH-8634 Hombrechtikon, Switzerland. Elevenlarch- and pine-feeding populations of the larch budmoth were studied in Western Europe using 24 allozyme loci. Hierarchical F-statistics between larch and pine forms (FSTO.O65) were much greater than those between localities within larch (FSTO.0O2) and within pine (FST 0.026), even when sympatric host-associated populations were considered. Analysis of Nei's genetic distance produced similar results, and a UPGMA tree of all populations consistently clustered samples from the same hosts together. Three loci, autosomal Pgm and Mdh-s and sex-linked Idh-s, were the most important loci involved in differences between the two forms. Previously considered to be conspecific, it is now clear that the larch and pine biotypes of Z. diniana are either host races (sensu Diehi & Bush, 1984) or good sympatric species which hybridize rarely. Regardless of taxonomic status, the distribution of larch and pine in the Pleistocene, phenological differences between the two hosts, together with the great vagility of Z. diniana, an outbreak pest, suggest that sympatric or parapatric differentiation is at least as likely as allopatric divergence. Keywords: allozymes,host races, populationstructure, sex linkage, speciation, Zeiraphera diniana Introduction host-associated species which rarely, if ever, hybrid- ize. Whether or not these host races initially evolved Whethersympatric speciation is common is one of in allopatry, the existence of hybridizing host races the more controversial issues in biology (Mayr, 1970; implies that there can be a stable route for sympatric Grant & Grant, 1979; Futuyma, 1983; Diehl & Bush, speciation because gene flow is not able to swamp 1984; Bush, 1994). By sympatric speciation we mean genetic differentiation. a process of genetic differentiation into species Host-associated forms have been found in many occurring between populations that are within insect species (Diehi & Bush, 1984); however, the normal dispersal distance of each other. The wide- best-studied examples often either lack significant spread belief in the alternative, allopatric model is genetic differentiation (Mitter & Futuyma, 1979; so prevalent perhaps because it is simpler to think of Jaenike & Selander, 1980; Menken, 1981) or are divergence in the absence of gene flow, rather than sibling species with no known gene exchange (Gutt- because of any theoretical implausibility of diver- man et a!., 1981; Pashley, 1988). One of the few gence in sympatry (Barton & Charlesworth, 1984). cases that stands up to close examination is that of 'Host races' (sensu Diehi & Bush, 1984), where host- the apple versus hawthorn races of Rhagoletis porno- associated differentiation of phytophagous or para- nella (Feder et a!., 1988, 1994; McPheron et a!., sitic insects is maintained in the face of gene flow, 1988; Smith, 1988; Bush et al., 1989). Overall, the provide an important intermediate link between host scarcity of 'true' host races, as defined above, casts adaptation within panmictic populations and 'good' doubt on the idea that sympatric speciation is com- mon. The lack of examples of true host races could, however, be because few likely cases have been *Correspondence studied in much detail. This is the first study to use 416 1995The Genetical Society of Great Britain. HOST RACE FORMATION IN ZE/RAPHERA 417 genetic markers to investigate the host-associated data). (v) Pheromones: larch and pine forms differ populations of Zeiraphera diniana. in female pheromone blend (Guerin et a!., 1984). The larch budmoth, Zeiraphera diniana Guenée is In spite of these differences, the two forms have a classic outbreak pest of coniferous forests across always been considered members of the same the Palaearctic (Varley et al., 1973; Berryman, 1986). species because: (i) they interbreed freely in the The species consists of host-associated forms feeding laboratory, with no distortion of sex ratio in the on larch (Larix), pine (Pinus) and spruce (Picea). progeny (Table 1; Baltensweiler, 1977); (ii) studies The population biology of the larch budmoth has of variation in female pheromone blend, variable been studied in great detail for over 30 years by response of males to these pheromones and the Baltensweiler et al. (reviewed by Baltensweiler et a!., intermediacy of F1 hybrids suggests that introgres- 1977; Baltensweiler, 1993). This moth has highly sion should be common in the wild (Priesner, 1979; cyclical population dynamics above 1000 m in larch Day, 1984; Guerin et al., 1984; Baltensweiler & forests of the Alps: every 8—10 years there are out- Priesner, 1988); (iii) their populations fluctuate in breaksduring which populations peak at synchrony; and (iv) probable hybrid forms are often 100,000-fold greater densities than during the subse- found on larch (Baltensweiler et al., 1977; Baltens- quent crash. Associated with this population fluctu- weiler, 1993). ation is an apparent genetic cycle involving 5th We are currently engaged in an investigation of instar larval colour morphs; dark morphs predom- genetic differences between larch and pine forms of inate during population build-up, whereas pale Z.dinianaand the potential relationship between morphs become commoner during the decline genetics and population dynamics. In this paper, we phase. The differences between the dark and pale use allozymes to investigate genetic differences forms of larvae on larch are similar to the differ- between host-associated biotypes. Only one previous ences between larvae on larch and pine; on larch, study on the allozymes of Z. diniana exists, in which larvae are predominately dark, on pine they are only three loci were investigated and the differences pale. This suggests that host-associated differentia- between host races were not studied (Verdon, 1978). tion contributes to the cycles of genetic variation on larch, and may even be involved in population regu- Materialsand methods lation (Baltensweiler, 1993). There is a variety of differences between the host- Atotal of 300 larvae were collected during summer associated populations of Z. diniana, apart from 1994 in 11 larch- or pine-feeding populations at nine larval colour pattern. (i) Size: larch and pine form sites (Fig. 1 and Table 2). About 80 per cent of adults are not known to differ in structure, but larch these larvae were frozen in liquid N2. Others were moths are on average larger (W. Baltensweiler, transported to the laboratory and reared to adult- unpublished data). (ii) Phenology: egg hatch of the hood. These adults were used to test for the possible larch form is synchronized with the flush of larch effects of host plants, parasitoids or pathogens, foliage, about 2—4 weeks earlier than the pine form which might have affected larval electrophoretic and its host shoots (Day, 1984). (iii) Host prefer- banding patterns. Adults could be sexed in the usual ence: larch and pine forms tend to lay eggs on their way. We also found that males of later instar pine own hosts (Maksymov, 1959; Bovey & Maksymov, form larvae could be detected because their dark 1959). (iv) Host-associated larval survival: the larch testes were visible through the semi-transparent cuti- form does not normally survive either on pine or cle. Sexing was more difficult for last instar larch spruce, whereas the pine form does well on all three form larvae, nearly all of which have a heavily pig- hosts (Day, 1984; W. Baltensweiler, unpublished mented cuticle. Table 1 Sex ratio in larch budmoth hybridst Type of cross No. of pairs F1 females F1 males Females (%) P (test: no distortion) L x L 23 621 628 49.7 >0.5 PxP 11 117 128 47.8 >0.5 LxP 5 211 198 51.6 >0.5 PxL 4 98 82 54.4 >0.1 tBroods are as mentioned in Baltensweiler (1993). L: larch form; P: pine form. The Genetical Society of Great Britain, Heredity, 75, 416—424. 418 I. EMELIANOV ETAL. Fig. 1 Geographical locations of 11 samples of Zeiraphera diniana. Abbreviated geographical names used further in this paper (Table 2, Fig. 2) are given in bold. BRITAIN: 1: North Yorkshire (NY), two individuals from two sites, Harwood Dale Forest (grid ref. NZ960988) and Cropton Forest (NZ810937), larvae on Japanese larch (Larix leptolepis); 2: Brachour Forest (Brac, DN090500), on Pinus contorta; 3: North Dalchork Forest (Dat, NC550200), on P contorta. FRANCE: 4: Méribel (Mér), 45°23'N 6°35'E, on P cembra; 5: Bois Les Combes (LC), near Briançon, 44°54'N 6°34'E, on L. decidua; 6, 7: Bois les Ayes sites 1 and 2 (LA! and LA2), near Briancon, 44°50'N6°39'Eand 44°51'N 6°41'E, respectively, larvae on both P cembra and L. decidua at each site; 8: Mont- génèvres (Mont), near Briançon, 44°56' N 6°43'E, larvae on L. decidua. SWITZERLAND: 9: Enga- din Valley (Eng), 46°26' N 9°50'E, on L. decidua. Larvae were cut in half and adults were also split where (Richardson et a!., 1986; Mallet et al., 1993) between the thorax and abdomen, after trimming with minor modifications. legs and wings. Posterior larval segments and thora- A total of 19 enzyme systems of 36 tested were ces of adults were kept for DNA extraction; anterior sufficiently scorable and active, and were selected larval sections and adult abdomens were homoge- for further analysis (abbreviated name of enzyme, nized on ice in 25—35 izL of extraction buffer con- EC no., and buffer system are shown in paren- taining 0.01 per cent NADP, 0.1 per cent theses): glutamate-oxaloacetate transaminase (Got fJ-mercaptoethanol and 0.1 per cent Triton X-100 2.6.1.1, TGB), phosphoglucose isomerase (Pgi detergent (Richardson et al., 1986) and centrifuged.