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
Volume 29 1975 Number 2
PAPILIO "GOTHICA" AND THE PHENOTYPIC PLASTICITY OF P. ZELICAON (PAPILIONIDAE)
Arthur M. Shapiro Department of Zoology, University of California, Davis, California 95616
Since the description of Papilio gothica Remington (Remington, 1968), its taxonomic status and the nature of its relationship with P. zelicaon Lucas have remained unclear (Clarke & Sheppard, 1970). Papilio gothica (type locality Gothic, Gunnison Co., Colorado) was proposed to apply to univoltine, montane populations from the Rocky Mountains and adjacent areas, showing an alleged host preference for (or restriction to) Pseudocymopterus montanus (Gray) Coulter & Rose (Umbelliferae), and differentiable from lowland California P. zelicaon in the phenotypes of their Fi hybrids with P. polyxenes Fabricius and P. bairdii Edwards. Actual or potential reproductive isolation from P. zelicaon was not demonstrated. Remington was unable to find any "all or none characters ... by which every dead specimen would be infallibly recognized," but he did provide a list of 11 characters whereby "most gothica differ from most zelicaon. . . ." He cited four of these as "the most reliable for distinguishing adults of montane Colorado gothica from lowland coastal Cali-fornian zelicaon" In his discussion of the geographical distributions of zelicaon and gothica he tentatively assigns genetically unknown populations to one or the other name, based on these pattern characters.
In many species of butterflies lowland, multivoltine populations produce vernal (or vernal/autumnal) phenotypes which are very similar to—if not indistinguishable from—those emerging in the single annual brood at high elevations or latitudes. Klots (1951) refers to this situation in Papilio glaucus L. ("canadensis" phenotype) and Lycaenopsis pseu-dargiolus Bdv. and LeC. ("lucia"). In California, it occurs in Polites sabuleti Bdv. ("tecumseh" phenotype) and Pieris occidentalis Reak. ("calyce"), to name two species under study in my laboratory. Papilio
80
Journal of the Lepidopterists' Society
zelicaon is continuously brooded in the Sacramento Valley of California, flying from March-October or November, with four to five broods. Since virtually every multivoltine butterfly studied (even the dull, dark species of Erynnis; Burns, 1964) shows seasonal phenotypic changes, it is scarcely surprising that P. zelicaon does so. The gestalt of the spring phenotype is immediately recognizable, but it has never been described in detail, or quantified as frequencies and correlations of individual character states. I propose that the seasonal phenotypic variability of P. zelicaon has a bearing on the validity of Remington's characterization of P. gothica.
When Remington's set of 11 gothica characters is examined in large samples of lowland California zelicaon, it becomes evident that spring zelicaon differ from their summer counterparts in precisely the same ways that gothica is alleged to differ from zelicaon. I have examined 320 zelicaon collected below 500' in six central California counties in 1972, 1973, and 1974, mostly as singletons or in series of ten or fewer. There are, however, two large series collected at single localities on single days: 12 3, 169 from Suisun City, Solano Co., 20 March 1974, and 353, 4$ from the American River Parkway, Sacramento Co., 2 June 1973. These represent the first and second generations in their respective localities and years. Suisun City (elevation ca. 10') and the American River (ca. 15') are approximately 50 air miles apart, and potentially isolated from each other by the Inner Coast Ranges, which rise from a low pass at Vacaville, eight miles east of Suisun, to a crest of 2800' a few miles to the north. However, hilltopping P. zelicaon have been taken along the crest, and breeding populations exist in and on both sides of the pass, along Interstate Highway 80. Suisun City has a distinctly more maritime climate than the American River; it is exposed to afternoon sea breezes and occasional summer fog entering through the Carquinez straits. Both populations feed on the introduced umbelliferous weed, Foeniculum vulgare Mill., to the apparent exclusion of all other (native and introduced) umbellifers.
The four "most reliable" gothica characters (Remington's characters 2-5) are:
(2) Forewing below, in postmedian broad yellow band, with the anterior spot having an outer edge strongly offset from a line drawn through the outer edges of spots 2-9.
(3) Forewing below with postmedian spot 2 tending to have its outer edge forming an angle with its caudal edge only slightly greater than 90°. (In zelicaon this is described as "much greater than 90°.")
(4) Forewing above, near coastal edge, usually with 2 fine yellow lines opposite end of cell, one anterad and one posterad of the closely parallel veins Sc and Ri.
(5) Hindwing above has cell Cu2, near the anal margin, with basal dark
Volume 29, Number 2
81
Table 1. Distribution of character states in two samples of lowland California
Papilio zelicaon. The characters are described in the text (G = gothica state; Z = zelicaon state).
Remington's
numbered characters 2 3 4 5
Suisun City, 20 March 74 12^^ 8G:4Z 4G:8Z 6G:6Z 7G:5Z
(1st generation) 16$$ 10G:6Z 11G:5Z 13G:3Z 8G:8Z
|
39 |
8G:31Z |
7G:32Z |
4G.-35Z |
10G:29Z |
|
x2 P: |
12.65 « .005 |
9.9 <.005 |
21.93 « .005 |
6.6 -.010 |
28 18G:10Z 15G:13Z 19G: 9Z 15G:13Z
American River, 2 June 73 35 S $ 6G:29Z 6G:29Z 4G:31Z 7G:28Z
(2nd generation) 4$$ 2G: 2Z 1G: 3Z 0G: 4Z 3G: 1Z
Xi2 within characters
color usually extending far distad of origin of vein Cu2 from cell. (Described as greatly reduced in zelicaon males.)
For each specimen in the two long one-day series, each of these characters was scored as being in the gothica (G) or zelicaon (Z) state. The results are summarized in Table 1, along with the chi-square test for homogeneity of the two samples in each character; for all four characters the distributions of states in the two samples differ at a high level of significance. As may be seen from Table 2, the frequency of gothica character states deviates significantly from a random (Poisson) pattern in the spring, but not the summer sample. Point correlation coefficients (Poole, 1974) were calculated within each sample between all possible pairs of characters to test for association among character states (Table 3). Characters 3 and 4 have the strongest association in both samples, while 3 and 5 are essentially independent of each other. It is clear that the spring phenotype is not produced in toto, by the action of a major "switch-gene."
Remington suggests that gothica-zelicaon sympatry should be sought "at mountain-lowland contacts in central and eastern California/' Can the distribution of gothica characters in these samples be explained in terms of introgressive hybridization? The American River Parkway sits at the base of one of the major canyons of the west slope of the Sierra Nevada, precisely where Remington predicts contact between zelicaon and gothica, but the June sample from there shows essentially no gothica tendencies. Suisun City is very nearly on the coast and is within easy flying range of the large vacant-lot populations in Oakland and Richmond. The March sample from Suisun City is phenotypically strongly reminiscent of the picture Remington portrays of gothica. These ob-
82
Journal of the Lepidopterists' Society
Table 2. Distribution of gothica (G) and zelicaon (Z) character states among numbers of characters, and their fit to a Poisson distribution (assumed independence among the characters).
Number of Number of
characters Individuals Number Expected (O- E)2
in G state Observed (O) from Poisson (E) E
Suisun City, 20 March 74
0 0 7.84 7.84
1 4 9.80 3.43
2 11 6.44 3.23
3 11 4.08 2.77
4 2 0.84 1.60
xa=18^
p « .005 American River, 2 June 73
0 19 11.65 4.64
1 12 14.04 0.29
2 6 9.36 1.21
3 2 1.56 0.12
4 0 0.39 0.39
*2 = a65~ .10 > p > .05
servations make no sense biogeographically, but they do seasonally. I have 45 summer zelicaon (collected on 16 different days) from Suisun and 20 spring zelicaon (6 days) from the American River, and their phenotypes confirm the seasonal pattern. (For example, of six taken on 19 March 1974 at the American River, one has 1, four have 2, and one has all 4 of its characters in the gothica state.) The only alternative hypothesis, that gothica and zelicaon are sympatric in lowland California with the former being spring-univoltine and the latter summer-multi-voltine, is not only not parsimonious; it is demonstrably incorrect from my breeding experiments at Davis.
The tendency for spring zelicaon to resemble gothica is not restricted to the four "most reliable" characters. In the Suisun and American River samples, nine of the eleven characters are significantly more like gothica in spring. The tenth, ground color of females, could not be evaluated because so few American River females were on hand. The eleventh, absence of yellow striping in the ventral forewing discal cell, was proposed as a trait distinguishing zelicaon and gothica from P. brucei Edwards.
Volume 29, Number 2
83
|
2 |
-.096 |
-.024 |
-.344 |
|
3 4 |
+ .126 |
0 -.192 |
|
|
3 |
4 |
5 |
|
|
2 3 4 |
+ .258 |
+ .247 +.282 |
+.105 + .004 -.212 |
Table 3. Point correlation coefficients (V) between all possible pairs of characters 2, 3, 4 and 5. The variances have been omitted.
Remington's numbered characters 3 4 5
Suisun City, 20 March 74
American River, 2 June 73
None of this proves anything about the specific distinctness of gothica. If speciation is to be defined biologically, it can be "proven" only by the demonstration of reproductive isolation, a demonstration Remington has not made (Clarke & Sheppard, 1970). Decisions on species distinctness based entirely on statistical analysis of dead specimens, in the absence of biological data, can readily lead to absurd conclusions, e.g., the revision of Pieris, by Abbott et al. (1960) in which four very distinct (and widely sympatric) species group "protodice" species are synonymized. On the other hand, there is no doubt that sibling species exist which are indistinguishable to the museum taxonomist but meet all the biological criteria of speciation. Ayala (1973) described two partially isolated subspecies of the Drosophila willistoni group which are "nameable" only statistically, by the frequency pattern of isozymes when they are subjected to starch gel electrophoresis! If Papilio gothica proves to be a valid species, it will be "nameable" only by the locality label. Remington's characters are clearly useless for discriminating individual gothica from individual spring specimens of multivoltine zelicaon. (Since he did not quantify his diagnosis, we do not even know whether series of Colorado gothica and spring California zelicaon are statistically separable.) The only readily recognizable entity is summer, lowland zelicaon. The adaptive significance of these characters is unknown (the blackening in cell Cu2 is likely to function in thermoregulation), but since they seem to be associated with diapause in both uni- and multivoltine populations, there is no reason to predict any correlation with the gothica X polyxenes (or bairdii) hybrid phenotype—on which the name gothica now wholly depends.
Literature Cited
Abbott, W. P., L. S. Dillon & R. K. Shrode. 1960. Geographic variation in Pieris protodice Boisduval and LeConte. Wasmann J. Biol. 18: 103-127.
84
Journal of the Lepidopterists' Society
Ayala, F. J. 1973. Two new subspecies of the Drosophila willistoni group. Pan-Pacific Entomol. 49: 274-279.
Burns, J. M. 1964. Evolution in skipper butterflies of the genus Erynnis. Univ. Calif. Publ. in Entomol. 37: 1-214.
Clarke, C. A. & P. M. Shepparjd. 1970. Is Papilio gothica (Papilionidae) a good species? J. Lepid. Soc. 24: 229-233.
Klots, A. B. 1951. A Field Guide to the Butterflies. Houghton Mifflin, Boston. 349 p.
Poole, R. W. 1974. An Introduction to Quantitative Ecology. McGraw-Hill, New York. 532 p.
Remington, C. L. 1968. A new sibling Papilio from the Rocky Mountains, with genetic and biological notes. Postilla 119: 1-40.
CORRECTIONS TO TWO ARTICLES ON ABERRANT CYNTHIA (NYMPHALIDAE)
Two recent articles on aberrant Cynthia have contained errors of identification which should be corrected to prevent perpetuation of inaccurate records. Phillips (1971, Great Basin Natural. 31: 256-260) reported the capture of Cynthia cardui ab. elymi Rambur from Salt Lake City, Utah. This aberration, illustrated in color, is clearly C. annahella ab. letcheri Grinnell, which is easily distinguished from C. cardui ab. elymi by the forewing discal cell marking which completely crosses the cell in C. annahella but remains a spot against the anterior cell margin in C. cardui, and by the forewing apex which is pointed at Mi in C. annahella and more rounded in C. cardui. C. cardui ab. elymi, then, remains unrecorded for Utah.
Shapiro (1973, Pan-Pac. Entomol. 49: 289-293) described and illustrated aberrations of C. annahella. His Figs. 1 and 2 are indeed C. annahella, but Fig. 3 is C. cardui showing tendencies toward ab. elymi. The same characteristics used to separate C. annahella ab. letcheri and C. cardui ab. elymi in the previous article can be used to correctly determine this specimen.
When used with care the excellent color illustrations of aberrant Cynthia (= Vanessa) in Comstock's Butterflies of California (1927) should leave no doubt as to the correct species identifications of some aberrations. However, the above mistaken identifications show the difficulty of applying names to aberrations, which by their very nature can be quite variable.
Thomas E. Dimock, 111 Stevens Circle, Ventura, California 93003.