Click here for the original journal page (in Acrobat pdf format).

The text below is grayed out because it is not intended to be read. It is a necessarily imperfect OCR of the original and is only used by a search engine.

Journal of the Lepidopterists' Society 43(3), 1989, 217-228

GENETIC DIFFERENTIATION AMONG CALIFORNIA

POPULATIONS OF THE ANISE SWALLOWTAIL

BUTTERFLY, PAPILIO ZELICAON LUCAS

Mark L. Tong1 and Arthur M. Shapiro2

Department of Zoology, University of California, Davis, California 95616

ABSTRACT. The anise swallowtail butterfly, Papilio zelicaon Lucas, is widely distributed in California. California zelicaon are composed of low- and high-elevation eco-types defined by host-plant preference and diapause physiology. Electrophoretic-genetic surveys of 14 loci over 10 populations (157 samples total) demonstrate great similarity among these ecotypes, suggesting that their adaptive differences may be defined by a small number of loci rather than broad genomic differentiation.

Additional key words: ecotypes, electrophoresis, Papilionidae.

The anise swallowtail butterfly, Papilio zelicaon Lucas, is native to western North America, where it is widely distributed (Tyler 1975), In central California, zelicaon is found in a wide variety of habitats from sea level to tree line (Table 1, Fig. 1). Populations at the same latitude exhibit diapause phenologies from univoltine (one generation/yr) to multivoltine (up to four/yr) as a function of habitat elevation, length of growing season, and larval host plant (Sims 1979).

Populations in the Coast Range and the Sierra Nevada above 400 m primarily utilize native Umbelliferae including Lomatium, Angelica, and Cymopterus. These native plants are available to zelicaon larvae from spring to midsummer when the onset of hot, dry weather renders the leaves too hard and dry for the larvae to ingest. These populations are univoltine, though in the montane Sierra a second generation occasionally occurs (Sims 1979, Shapiro unpubl.).

Lowland populations (below 400 m) today feed almost exclusively on sweet fennel (Foeniculum vulgare Miller, Umbelliferae), which is common throughout coastal and interior lowland California (Munz 1970), and also on orange (Citrus sinensis Osbeck, Rutaceae) which has been grown commercially in California since 1841 (Opitz & Piatt 1969). Both plants are available to zelicaon 8-12 months per year, allowing these populations to breed continuously (Sims 1983). Fennel and orange were introduced to California by Spanish missionaries in the 18th Century (Hutchinson 1969). Both produce natural compounds similar to those in native Umbelliferae (Dethier 1941). Before this zelicaon was presumably univoltine, being limited by ephemeral host plants at low elevations and the short growing season in the mountains (Sims 1983). The introduction of these perennial host plants probably

1 Current address: 109 Berwick Drive, Pittsburgh, Pennsylvania 15215.

2 To whom reprint requests should be sent.

218

Journal of the Lepidopterists' Society

enabled zelicaon to "switch" its ovipositional preference to the introduced plants or to disperse to areas where only the introduced species were available, or both. This, in turn, allowed multivoltinism to evolve in the lowland areas where these plants are abundant (Shapiro & Masuda 1980, Sims 1983).

Papilio zelicaon populations are consistent with the ecotype concept first proposed by Turesson (1922) for hawk weed (Hieracium, Astera-ceae): plants from different habitats were shown to be phenotypically distinct even when grown under identical conditions, thus demonstrating a genetic basis for the differences.

Using wild zelicaon and a laboratory strain selected for nondiapause, Sims (1979) demonstrated that univoltine zelicaon populations have significantly higher diapause incidence (photophase required to induce diapause) and intensity (duration of chilling needed to terminate diapause) than multivoltines, and that this phenomenon is genetically based. Sims (1983) showed that incidence and intensity are polygenically inherited, with intensity being affected by maternal phenotype. These characteristics are maintained under varied environmental regimes (Sims 1979, Shapiro unpubl.).

The present study began as an attempt to use electrophoretic analysis to determine whether orange-feeding zelicaon in northern California evolved independently of orange-feeding zelicaon in southern California, or had been introduced inadvertently from the south. Papilio zelicaon was reported as an orange pest as early as 1909 near Visalia, Tulare Co. (Coolidge 1910), and in the 1960's near Chico, Butte Co. (Shapiro unpubl.). This study also investigates the degree to which differentiation of zelicaon into low- and high-elevation diapause ecotypes is reflected by electrophoretically detectable genetic variation.

Materials and Methods

Electrophoresis, a commonly used method in biochemical system-atics, is based on the movement of charged particles under the influence of an electrical field (Ferguson 1980). Proteins carry a net electrical charge depending on amino acid structures and environmental pH. The rate at which proteins migrate through a support medium is related to their size and shape and is proportional to net charge. Different proteins with different electrophoretic properties migrate at different rates under identical test conditions.

Differential migration of homologous proteins is detectable and of special interest in biochemical systematics. Such differentiation is presumed to reflect differences in nucleic acid sequences that encode proteins. The degree of electrophoretically detected differentiation is thought to reflect the extent of evolutionary divergence between sam-

Volume 43, Number 3

219

1       Gazelle

2       Washington

3       Hemet

4       Orland

5       Butts Canyon

6       Suisun

7      Blue Ridge

8       Rancho Cordova

9      Castle Peak 10       Auburn

Fig. 1. Location of populations studied.

pled taxonomic groups, although the structural proteins studied represent only one segment of the overall genome.

Adult zelicaon were collected during 1984 and 1985 flight seasons at the sites in Table 1. Captured specimens were frozen live and stored at — 70°C to prevent protein denaturation.

In preparation for analysis, thoraces were excised and homogenized in 600 Ail of glass-distilled water with a Teflon-coated tissue grinder. Homogenate was absorbed onto 2x9 mm wicks of #3 Whatman paper and applied to the gels. Horizontal slab gels were made with Sigma starch and were prepared and run for 5 h using methods described in Ayala et al. (1972, 1974a).

After running, gels were cut into four 2-mm-thick slices so that each

Table 1. California field site characteristics (Shap

Volume 43, Number 3

221

Table 2. Enzymes assayed.

* REG: Gel buffer—9 mM Tris, 3 mM citric acid, pH 7.0. Electrode buffer—135 mM Tris, 45 mM citric acid. JRP: Gel buffer—76 mM Tris, 5 mM citric acid, pH 8.65. Electrode buffer—300 mM boric acid, 60 mM NaOH. DH: Gel and electrode buffer—8.7 mM Tris, 8.7 boric acid, 1 mM EDTA, 1 mM /3-NAD+, pH 9.0.

sample was tested for four enzymes. Table 2 lists the enzymes assayed. Specific staining systems and gel fixation techniques are described in Ayala et al. (1972, 1974a).

Fixed gels were scored after each run using a light box. Loci were characterized and interpreted as for Pieridae, for which the genetic basis of the electrophoretic banding patterns has been demonstrated in an extensive breeding program (Geiger 1981, Burns & Johnson 1971). Electromorphs were recorded as distance (mm) migrated from the origin.

Electromorph frequencies (considered as allelic frequencies) were used to calculate I, a statistic of genetic identity between taxa (Nei 1972), for all pairwise comparisons of populations. I-values were analyzed using the UPGMA method of cluster analysis (Ferguson 1980).

G-tests (Sokal & Rohlf 1981) were performed on genotype frequencies in the populations represented by large samples (>14 individuals) to determine whether observed frequencies for each population were consistent with Hardy-Weinberg equilibrium, and whether all populations can be considered to represent a single panmictic population.

Results

Table 3 shows the electromorph frequencies for each population. Of the 14 loci assayed, three are polymorphic: PGI, PGM and MPI.

Results of G-tests are displayed in Table 4. Most loci exhibit Hardy-Weinberg equilibrium. However, weighted-average results show that among the populations examined, zelicaon does not exhibit Hardy-Weinberg equilibrium and cannot be considered a single, panmictic population. Genotype frequencies are shown in Table 7.

222

Journal of the Lepidopterists' Society

-£>°°

PS (J

J2 o^--v

l«<3~

B2

O

^^D                       (NCOh CO CO tP (N ^D IO

^HOOOOOO<NcOcO^HOOO<NiO^OOOOOO

OdHHHHOCiddHddCidCiHHHHHH

IO CO <D IO 00000^(NTf^O

o o

^^DOOOOOO

iOOOO-hOOOOO.

00

NN                       CO 00 IO CO (N00ONO

^h00OOOO<N(NC0^hOOOc0I000OOOOOO

ho5                  a> 10 co co

H00OOOO(NC0NOO

dd-H-5-;-5dddd^©©ddd

05 CO 00 Ot^-nOOOOOO

io io co r*-

ooooocqoqco-no

o o

tF^D OOOOOO

iOOOO-hOOOOO-

CO^DOO^       05 05 00 ^

OOOOOcq^-n^OOO^H^D OOOOOO

lOOOO^OOOOO-

OO                       0000IO00                  COOb

(NOOOOOO^h^h^^hO             ^tlOOOOOOOO

dd^-i-i-i-idddd-HOOddd-H-5-i-i-5-H

IOIO                       ONOC0 ^D ^D O O ^D

OO^OOOO-h^^OOOOCnI^OOOOOOO

OO-h-h-h-hOOOO-hOOOOO-

^> ^h cq ^h                  onco

ooooooqoqcooio          coio^oooooo

iOOOO-hOOOOO-

o o o _ ooooo oq^^o

o o

IOIOOOOOOO

oddd—iooodd-l^H-H-i-5-1

o o

OOOOOIOIO             O

o^-i^-i^-i^^ddoo-HOOodd.

C0NOC0NOMOC0O(N^ONC000ONN^M^ ^h h(N(MtPC0N(N(Mh^t}<c0C0MN                  (N^h

~ 8§2^s        ±~

o jqO^o p^ pm <: s o £ cm fe^

hQQ

E.-M

' Ph ^ W h h

^oolww

Volume 43, Number 3

223

Table 4. G-test of genotype frequencies in populations >14.

, OBS „ * P < 0.05. G = 22 OBS In——; df = EXP

y2(n2 - n) (Ferguson 1980).

Table 5 shows the I-values calculated from electromorph frequencies for each pairwise comparison of populations. The average I-value is 0.980 ± 0.139, indicating a very high level of genetic similarity.

To highlight differentiation without altering phenetic clustering of populations, I-values were recalculated using only data for the three polymorphic loci (Table 6). By excluding the background of mono-morphic loci, a more useful graphic analysis can be generated. The UPGMA dendrogram derived from these data is shown in Fig. 2.

Discussion

The high identity values (Table 5) are consistent with similar studies on other insects. Ayala et al. (1974b) reported I of 0.970 ± 0.006 for Drosophila willistoni populations sampled throughout Central and South America, substantiating previous conclusions about their relatedness based on reproductive relations. Brussard et al. (1985) surveyed genetic identity findings comparing 14 insect taxa (including 3 Lepidoptera), and reported I of 0.97 for local populations. It is likely that these values are conservative. Mutants of crucial glycolytic or catabolic enzymes are likely to be eliminated by selection (Bell 1976; Turner 1974; Zera et al. 1985).

Reproductive relations in zelicaon are not clear. Breeding trials among

Table 5. Genetic identity I-value matrix using 14 loci (I = 0.980 ± 0.139).

224                                                    Journal of the Lepidopterists' Society

Table 6. I-value matrix using only the three polymorphic (I = 0.867 ± 0.100) loci.

multivoltines from northern and southern California and the Central Valley show these populations to be intercompatible (Shapiro unpubl.). Alternatively, Sims (1983) suggests that univoltines and multivoltines are not fully intercompatible because of male-biased hybrid broods. However, control (within-population) data are not available in adequate numbers to validate this conclusion.

While I-values suggest that all zelicaon populations are conspecific, the weighted-average G-tests show that zelicaon is, of course, not pan-mictic over its entire range. Figure 2 suggests that populations can be clustered on the basis of geographic proximity.

Gazelle (Shasta Valley) is most genetically dissimilar, and is probably more geographically isolated as well. Washington and Auburn are Sier-ran west slope univoltines. Orland, Suisun and Rancho Cordova are Central Valley multivoltines. Castle Peak, Butts Canyon, Blue Ridge, and Hemet represent univoltine and multivoltine populations in a mixture of very diverse ecological contexts.

The germinal issue is to what degree populations are reproductively isolated by host-plant selection and physical distance. Certainly, the breeding trials and high I-values suggest that all zelicaon populations are potentially intercompatible. However, voltinism may be a genetically heritable trait that divides zelicaon into low- and high-elevation ecotypes (Clarke & Sheppard 1970).

Our data support Sims's (1983) contention that zelicaon diapause physiology and host-plant selection are highly plastic. The clustering of the orange-feeding Orland population with other Central Valley populations rather than with Hemet implies that the northern and southern orange-feeders evolved separately. While fennel is abundant in lowland areas, and is heavily used (Shapiro 1974a, 1974b), the use of orange may allow zelicaon to increase its range despite the inferiority of orange as a host plant (Masuda 1981).

Volume 43, Number 3

225

Table 7. Observed genotype frequencies for the three polymorphic loci in populations where n > 14.

Butts Canyon (North Coast Range serpentine) and Castle Peak (Sier-ran volcanic alpine) probably represent relict zelicaon populating rocky, unforested environments with endemic host plants. Other Lepidoptera are known to be similarly disjointly distributed between the Coast Range serpentines and the alpine Sierra; Papilio indra, Pieris sisymbrii, and Euchloe hyantis all occur obligately in these areas with few or no intervening populations (Shapiro unpubl). Clustering of Blue Ridge (east of the Vaca Hills, the easternmost part of the Inner North Coast Ranges in Yolo and Solano cos.) with these postulated relict populations rather than with other multivoltines in the Central Valley is especially interesting. Papilio zelicaon was not seen in the Vaca Hills during summer in field studies initiated by Shapiro (unpubl.) in 1972. Males were seen on the ridge-tops, but only in spring coinciding with such behavior on Coast Range serpentines to the north. At this time, the site had one patch of 10 fennel plants. In 1975, females were observed

226

Journal of the Lepidopterists' Society

Blue Ridge

Castle Peak

Hemet

Butts Canyon Orland

Suisun

RanchoCordova Washi ngton

Auburn Gazelle

0.800

0.900

1.000

I -V A LUE Fig. 2. Phenogram of P. zelicaon populations (UPGMA; Ferguson 1980).

ovipositing on fennel. By 1978, fennel was spreading rapidly in disturbed areas and zelicaon showed evidence of four generations in one year. Presently, there are over 500 fennel plants along three miles of road in this area, and it continues to spread. It has been presumed that the multivoltine Vaca Hills zelicaon are upslope colonists from multi-voltine Central Valley populations. Our study suggests, rather, that they are at least partially downslope colonists from univoltine ridge-top (Coast Range) populations. If this is the case, they have very rapidly evolved multivoltinism, apparently as an adaptation to the spread of fennel. This supports the plasticity of host plant- and diapause-"switching" proposed by Sims (1983) to explain the evolution of multivoltinism. Certainly, zelicaon is physically capable of having colonized these canyons from the Coast Range. Shields (1967) demonstrated that zelicaon is a hilltopping species; adult males and receptive females congregate on summits to mate, thereby promoting gene flow among neighboring populations. Shields determined that adults are capable of traveling several km per day.

Studies by Ehrlich and Raven (1969) and Endler (1973) suggest that populations undergoing sufficiently strong divergent selection will differentiate despite the counter-effects of continuous gene flow. This has been observed in wild Lepidoptera with populations showing differ-

Volume 43, Number 3

227

entiation in metrical traits as a result of differential selection, despite close proximity and gene flow (Creed et al. 1959, Clarke & Sheppard 1962). If gene flow along the Coast Range ridge-tops has been continuous, the Vaca Hills population has not only become multivoltine within three years time, but has done so with constant influx of univoltines from the Coast Range. Multivoltinism may be evolving through hybridization, or through selection. Multivoltinism shortens generation time and should, all other factors being equal, be selectively advantageous.

Wright (1943a, 1943b) theorized that a continuously distributed species exposed to different conditions of selection would differentiate if subdivided into partially isolated "islands" separated through inbreeding or limited dispersal ability. Papilio zelicaon is certainly distributed throughout habitats with different selection conditions and appears to be sufficiently vagile to be essentially continuous in distribution throughout major portions of its range. More finely focussed studies of nongly-colytic enzymes and mark-release-recapture studies on movement would help to determine the size and location of hilltopping regions and the appropriateness of Wright's "island" models to zelicaon.

Papilio zelicaon is composed of low- and high-elevation ecotypes defined by host-plant preferences and diapause physiology. These traits may be determined by a relatively small number of loci that are under strong selection pressure and whose distribution is not reflected by electrophoretically accessible glycolytic enzyme loci, which show great genetic similarity among populations.

Acknowledgments

We thank F. J. Ayala for use of laboratory facilities, Hansjiirg Geiger and P. Ward for helping us interpret electrophoretic data, and John Emmel for the Hemet specimens; also Douglas Engfer for developing software to compute I-values. This study forms part of California Agricultural Experiment Station Project CA-D*-AZO-3593, "Host Switching by the Anise Swallowtail," A. M. Shapiro, Principal Investigator.

Literature Cited

Ayala, F. J., J. R. Powell, M. L. Tracey, C. A. Mourao & S. Perez-Salas. 1972. Enzyme variability in the Drosophila willistoni group. IV. Genie variation in natural populations of D. willistoni. Genetics 70:113-139.

---------- 1974a. Genetic variation in natural populations of five Drosophila species and

the hypothesis of the selective neutrality of protein polymorphisms. Genetics 77:343-384.

---------- 1974b. Genetic differentiation during the speciation process in Drosophila.

Evolution 28:576-592.

Bell, M. A. 1976. Evolution of phenotypic diversity in Gasterosteus aculeatus super-species on the Pacific Coast of North America. Syst. Zool. 25:211-227.

Brussard, P. F., P. R. Ehrlich, D. D. Murphy, B. A. Wilcox & J. Wright. 1985. Genetic distances and the taxonomy of checkerspot butterflies (Nymphalidae: Nym-phalinae). J. Kansas Entomol. Soc. 58:403-412.

228

Journal of the Lepidopterists' Society

Burns, J. M. & F. M. Johnson. 1971. Esterase polymorphism in the butterfly Hemiargus isola: Stability in a variable environment. Proc. Nat. Acad. Sci. 68:34-37.

Clarke, C. A. & P. M. Sheppard. 1962. Disruptive selection and its effect on a metrical character in the butterfly Papilio dardanus. Evolution 16:214-226.

---------- 1970. Is Papilio gothica a good species. J. Lepid. Soc. 24:230-233.

Coolidge, K. R. 1910. A California orange dog. Pomona College J. Entomol. 2:533-

534. Creed, E. R., W. H. Dowdeswell, E. B. Ford & K. G. McWhirter. 1959. Evolutionary studies on Maniola jurtina: The English mainland, 1956-57. Heredity 13:

363-391. DETHIER, V. G. 1941. Chemical factors determining the choice of food plants by Papilio

larvae. Am. Nat. 75:61-73. Ehrlich, P. R. & P. H. Raven. 1969. Differentiation of populations. Science 165:1228-

1232. Endler, J. A. 1973. Gene flow and population differentiation. Science 179:243-250. Ferguson, A. 1980. Biochemical systematics and evolution. Blackie, Glasgow. 194 pp. Geiger, H. J. 1981. Enzyme electrophoretic studies on the genetic relationships of

pierid butterflies (Lepidoptera: Pieridae) I. European taxa. J. Res. Lepid. 19:181-

195. Hutchinson, W. H. 1969. California: Two centuries of man, land and growth in the

Golden State. American West Publishing Co., Palo Alto, California. 351 pp. Masuda, K. K. 1981. The effects of two host plants, Foeniculum vulgare Mill, and

Citrus sinensis Osbeck, on adult oviposition, behavior and larval and pupal biology

of the anise swallowtail butterfly (Papilio zelicaon, Lucas) (Lepidoptera: Papilioni-

dae). M.S. Thesis, University of California, Davis. 82 pp. Munz, P. A. 1970. A California flora. Univ. of California Press, Berkeley. 1681 pp. Nei, M. 1972. Genetic distance between populations. Am. Nat. 106:283-292. Opitz, K. W. & R. G. Platt. 1969. Citrus growing in California. Calif. Agric. Exper.

Station Service. 59 pp. Shapiro, A. M. 1974a. The butterfly fauna of the Sacramento Valley, California. J. Res.

Lepid. 13:73-82, 115-122, 137-148.

---------- 1974b. Butterflies of the Suisun Marsh, California. J. Res. Lepid. 13:191-206.

---------- & K. K. Masuda. 1980. An opportunistic new citrus pest. Cal. Agric. 34:4-5.

Shields, O. 1967. Hilltopping: An ecological study of summit congregation behavior

of butterflies on a southern California hill. J. Res. Lepid. 6:69-178. Sims, S. R. 1979. The reproductive biology and diapause dynamics of Papilio zelicaon (Lepidoptera: Papilionidae). Ph.D. Dissertation, University of California, Davis. 87 pp. Diss. Abstr. Order No. DA7815518.

---------- 1983. Inheritance of diapause induction and intensity in Papilio zelicaon. Heredity 51:495-500.

Sokal, R. R. & F. J. Rohlf. 1981. Biometry: The principles and practice of statistics

in biological research. 2nd ed. Freeman, San Francisco. 859 pp. Turner, B. J. 1974. Genetic divergence of Death Valley puphsh species: Biochemical

versus morphological evidence. Evolution 28:281-294. TURESSON, G. 1922. The genetic response of the plant species to the habitat. Hereditas

3:211-350. Tyler, H. A. 1975. The swallowtail butterflies of North America. Naturegraph Pub.,

Healdsburg, California. 192 pp. Wright, S. 1943a. Isolation by distance. Genetics 28:114-138.

---------- 1943b. An analysis of local variability of flower color in Linanthus parryae.

Genetics 28:139-156.

Zera, A. J., R. K. Koehn & J. G. Hall. 1985. Allozymes and biochemical adaptation, pp. 633-674. In Kerkut, G. A. & L. I. Gilbert (eds.), Comprehensive insect physiology, biochemistry and pharmacology, Vol. 10. Pergamon Press, Elmsford, New York.

Received for publication 5 July 1988; accepted 26 May 1989.