PROC. ENTOMOL. SOC. WASH. 109(2), 2007, pp. 295-308 GENETIC DIFFERENTIATION BETWEEN THE SIBLING AND SYMPATRIC FLOWER-HEAD INFESTING TEPHRITIDS: THE POLYPHAGE, TRUPANEA NIGRICORNIS (COQUILLETT), AND THE NARROWLY OLIGOPHAGOUS, T. BISETOSA (COQUILLETT) (DIPTERA: TEPHRITIDAE) Khouzama M. Knio, Richard D. Goeden, and David H. Headrick (KMK) Biology Department, American University of Beirut, P.O. Box 11-0236, Riad El Sohl, Beirut, Lebanon (e-mail: [email protected]); (RDG) Department of Entomology, University of California, Riverside, CA 92521, U.S.A.; (DHH) Horticulture and Crop Science Department, California Polytechnic State University, San Luis Obispo, CA 93407, U.S.A. — Abstract. Trupanea nigricornis and T. bisetosa are sibling, cryptic species that occur in sympatry in southern California. Trupanea nigricornis is generalist, infesting at least 8 tribes in the Asteraceae while T. bisetosa is mainly a specialist on wild sunflowers. Although the two species are very similar in morphology and biology, genetic differentiation between them was possible using isozyme electrophoresis. Among 14 resolved loci, 5 were polymorphic. Significant allele frequency differences were found between the two species for PGI, PGM, ME, and EST-1. Moreover, a fixed allele difference for the locus ACPH was detected, indicating absence ofgene flow between T. nigricornis and T. bisetosa in the sampled sympatric populations. The absence of hybrids could be explained by the results of cross-mating studies, which indicated that the two species did mate in the laboratory, but produced few, if any, viable offspring. Hybrid inviability acted as a post-mating barrier reducing gene flow between sympatric populations of T. nigricornis and T. bisetosa. Key Words: Isoenzyme electrophoresis, cross-mating studies, sympatry, Tephritidae, Trupanea, post-mating isolating barriers Trupanea nigricornis (Coquillett) and be recognized by the shape of the Y- T. bisetosa (Coquillett) are flower head shaped apical marking on the wing. This infesting tephritids that occur in sympat- marking is usually thin and distinct in T. ry in southern California. They are nigricornis females and broad and short sibling and cryptic species as both their in T. bisetosa females (Cavender and adult and immature stages show great Goeden 1983, Foote et al. 1993). morphological similarities and are diffi- Resource utilization studies showed cult to separate (Foote et al. 1993, Knio that the larvae of both species exploited et al. 1996a). The males can only be the flowerheads oftheirhosts in a similar distinguished by the color of the third manner and they fed on a similar number antennal segment, which is brown in T. of achenes, taking into account the size nigricornis and yellow in T. bisetosa. oftheir host (Knio et al. 2001). However, Females are more difficult to separate ecologically, these two sympatric species based on anatomy. Most (ca. 75%) can showed major differences. Trupanea ni- 296 PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON griconiis behaves as a generalist, infest- Biological characteristics permitting the ing many Asteraceae hosts belonging to development of new host races in sym- 8 tribes while T. bisetosa behaves as patry include mating on the host plant, a narrowly oligophagous species infest- positive correlation between host and ing a few Asteraceae species belonging to mate selection, and genetic control of only one tribe, the Heliantheae (Goeden host selection (Bush 1975). It is not 1985, 1992). Moreover, T. bisetosa coxAd known whether T. nigricornis and T. be considered as a specialist mainly on bisetosa meet all the biological criteria wild sunflowers, Helicmthus annuus L., its proposed by models of sympatric speci- most common host in southern Califor- ation. In addition, it is not known nia (Cavender and Goeden 1983). whether these two species hybridize in Although the oviposition behavior nature since there are no geographical differed between females of T. nigricornis barriers to isolate them and their host and T. bisetosa, only subtle differences plants very often occur side by side in were detected in the courtship and southern California. mating behavior ofthe two species (Knio This study investigates, using isoen- et al. 1996b). No apparent physical and zyme electrophoresis, whether gene flow temporal barriers seem to keep these two occurs between sympatric populations of closely related species from meeting and T. nigricornis and T. bisetosa in southern mating in the field. The hosts of T. California and whether the two species nigricornis bloom for a short period (1- produce viable offspring when crossed in 2 months) during the fall or the spring; the laboratory. It is one of a series of while the main host of T. bisetosa, H. comparative studies intended to shed annuus, is in bloom throughout the year light on the attributes that enable T. in southern California, in the absence of nigricornis to be a successful generalist frost. Nevertheless, spatial separation and T. bisetosa a specialist on wild appears to be an important factor sunflowers. keeping these two species from meeting Materials and Methods in areas of sympatry as adults of both — species were observed to meet and mate Insect collections. The flies used in on their respective host plants, and the electrophoretic studies were reared from timing of courtship displays was differ- mature flower heads containing third ent. In the field, males of T. nigricornis instar larvae or puparia. The flower were observed to exhibit their courtship heads were placed in glass-topped, behavior in the mornings while those of sleeve, insectary cages (34x32x35 cm) T. bisetosa exhibited courtship display in at the University ofCalifornia, Riverside the afternoon (Knio et al. 1996b). at 60% RH and 12/12 (LD) photoperiod However, in the insectary, cross-matings from 0500-1700 h. After emergence, the occurred between the two species and flies fed on honey striped on the inner resulted in viable and fertile eggs (Ca- glass wall of the cage and water for 2- vender and Goeden 1983). 4 d, and then they were frozen and — Because T. nigricornis and T. bisetosa stored in liquid nitrogen at 180'C. are sympatric and taxonomically close Adults of T. bisetosa (N = 231) were species (Foote et al. 1993), but differ in reared from heads of wild sunflowers, their mode of herbivory and do not Helicmthus annuus; those of T. nigricornis overlap in their host range, it is possible were reared from heads of Encelia that they diverged from one common fcn-inosa Gray (N = 252) and, in one ancestor or that one species evolved from instance, from heads of E. virginensis A. the other following host race formation. Nelson (N = 55). All collections were VOLUME NUMBER 109, 2 297 made at locations of sympatry, i.e., 40 min. until the bands became distinct, where both wild sunflowers and Encelia then fixed in methanol:water:acetic acid spp. were growing in adjacent patches. (5:5:1) (Berlocher 1980). The alleles were The sites ofcollections were Casa Blanca designated according to their Rf value, and Lake Perris (Riverside Co.), CA; the distance migrated by the allele di- and, Wheeler Canyon (Inyo Co.), CA. vided by the distance migrated by the — Electrophoresis. Horizontal starch gel front or the marker dye. electrophoresis was conducted according Buffer systems and staining solu- — to the techniques described by Berlocher tions. Initial work consisted of resolv- (1980) and Pasteur et al. (1987). The gels ing 14 enzymes using different buffer contained 11.16% (48 g starch + 430 ml systems (Table 1). Five enzymes were buffer) potato starch (Sigma Chemical found to be polymorphic in T. nigricornis Co.). and T. bisetosa populations. These were Each fly was homogenized with 40 \x\ phosphoglucose mutase (PGM), phos- 0.1% Triton 100-X grinding buffer using phoglucose isomerase (PGI), acid phos- a motorized pestle. The homogenate was phatase (ACPH), malic enzyme (ME), mm absorbed onto a 10X4 wick (What- and esterases (EST). For this reason, the man #1 filter paper). Electrophoresis flies were routinely tested for these was stopped when the marker dye, enzymes. bromophenol blue (0.1% in water), Two of these enzymes, PGM and migrated about 9 cm toward the anode. ACPH, were run on aminopropyl mor- Following electrophoresis, the gel was pholine citrate/ tris-citrate-EDTA buffer, sliced into three to four (2 mm) layers. pH 8.3: gel buffer containing 6.8 g/1 EDTA Each slice was placed in a staining box, monohydrate citric acid, 0.3 g/1 incubated at 37°C in a staining solution (disodium salt), and aminopropyl mor- specific for a certain enzyme for 30- pholine to pH 8.5; and electrode buffer Table 1. Enzymes analyzed in Trupcmea nigricornis and T. bisetosa using different buffer systems. PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON 298 containing 6.4 g/1 monohydratecitric acid, other day. The flies were allowed to feed aminopropyl morpholine to pH 5.0, tris on the yeast hydrolysate diet for 18 d topH 8.3. (ClaytonandTretiak 1972). The until the females reached sexual maturity enzymes PGI, ME and EST were run on (unpublished data). On day 18 after tris-citrate (Poulik) buffer: gel buffer con- emergence, the mature flies were paired sisting of (0.076 M tris-0.005 M citric according to the assigned mate and acid) 9.21 g tris and 1.05 g monohydrate transferred to smaller cages for the citric acid/ 1 1 distilled water pH 8.7; anMd following treatments: electrode buffer consisting of (0.3 1. T. nigricornis females X T. hisetosa borate) 18.55 g boric acid and 2.4 g males (six pairs). psoHdi8u.2m (hSyedlraonxdeirde/et a1l.1 1d9i7st1i)l.leTdhewatterirs,- 2. fT.emanligersic(osrinxipsaimras)l.es X T. hisetosa c2.i5trhateat-ED1T2A0 bvuoffhesr awnads rcau.n 4fo0rmAca;. 3. Tn.isnimgarliecsorn(siisxfpeamiarsl)e.s X T. nigricor- w3hheraetas1,50thveolPtsouaHnkd bcau.ff5e0r mwAas. run for 4. Tm.alheisse(tsoisxapamirasl)e.s X T. hisetosa fe- The enzymes were stained following 5. Unmated T. nigricornis females (six the recipes described by Pasteur et al. females). (19S8t7at)isatnicds.S—haAwchain-dsqPuraarseatdes(t19w7a0s).used 6. Ufenmmaaletse)d. T. hisetosa females (six to determine ifany significant differences in the allele frequencies existed between Each pair of flies or single female was or within T. nigricornis and T. hisetosa placed in 850 ml clear-plastic cage fitted populations. It was also used to detect with a basal water reservoir and deviations from the Hardy-Weinberg a screened lid to allow ventilation. The equilibrium. Since some loci had several lid of each cage was striped with honey alleles with low frequencies, the alleles and contained a cotton strip dipped in were pooled when it was necessary to the yeast hydrolysate solution. Each cage ensure that no expected frequency was was provided with a bouquet of imma- less than one and that no more than 20% ture flower heads, the excised peduncles of the expected frequencies was less than of which were emerged in the water five (Zar 1984). — reservoir and held in place by an Cross-mating experiments. The flies absorbent cotton wick that also provided used in cross-mating experiments were a water source for the flies. Trupanea reared from puparia dissected out of nigricornis females were provided with E. field-collected flower heads, E. farinosa farinosa heads while those of T. hisetosa for T. nigricornis and H. annuiis for T. were provided with H. annuus heads. The hisetosa. flower head bouquet was replaced every After emergence, male and female flies other day for a period of 2 months of each species were separated and (May-June), and was collected from placed in different glass-topped, sleeved, stems that had been covered with a fine cages in the insectary. Tightly wrapped mesh cloth to protect them from ovipo- cotton strips (ca. 2X1 cm) dipped in sition in the field. a solution of sucrose-yeast hydrolysate- The caged flower heads were dissected water (7:4:10) (Tsiropoulos 1978) were under a stereomicroscope and the num- affixed with adhesive tape to the inner ber of eggs in each was recorded. The wall surface ofthe cage as a food source, eggs were placed on a filter paper in addition to stripes ofhoney and water. (Whatman #1) moistened with physio- The cotton strips were changed every logical saline in closed glass Petri dishes. VOLUME NUMBER 109, 2 299 Rf AA BB AB CC BC (+) 0.32 0.3 0.28 94.4 5.6 Genotype (%) T. bisetosa (44.9 46 6.4 0.9/ T. nigricornis Fig. 1. Zymogram and genotypes (%) found in Tnipanea nigricornis and T. bisetosa populations for malic enzyme (ME). The Petri dishes were placed in an ofmajor bands found per locus. Among incubator at 27°C and checked each 14 resolved loci, only acid phosphatase, day for eclosion. The eggs were consid- phosphoglucose isomerase, phosphoglu- ered infertile if they did not hatch and cose mutase, malic enzyme, and esterases remained white with no sign of embry- were polymorphic in T. nigricornis and onic development after 2 weeks, i.e., no T. bisetosa populations. The other tested sign of the cephalopharyngeal skeleton enzymes were monomorphic indicating (Cavender and Goeden 1983). On the genetic similarity between the two spe- other hand, the eggs were considered cies. fertile if they hatched into active first Among the polymorphic enzymes, ME instar larvae. The number of fertile and and EST were not always clearly re- infertile eggs was recorded and percent- solved and therefore were not included in age eclosion for each treatment was the intraspecific variation analysis. Malic calculated. Analysis ofvariance was used enzyme behaved as a monomer. Two ME to compare the mean numbers of eggs alleles for were detected in T. laid between the different crosses and bisetosa populations while three were their controls. detected for T. nigricornis populations (Fig. 1). Allele 0.32 was almost fixed for Results T. bisetosa populations with a frequency Isoenzyme electrophoresis was used to of0.94 (Table 2). There was a significant ME test whether gene flow occurs between difference in the allele frequencies of sympatric populations of T. nigricornis between the two species (X^ = 89.2; df = and T. bisetosa. Table 1 summarizes the 2; p < 0.001). In the esterases, two loci isozyme loci resolved in this study, the were detected. EST-1 was polymorphic number of flies analyzed for each locus, with three common alleles in the two the buffer systems used and the number species (Fig. 2); significant allele fre- PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON 300 Table2. AllelefrequenciesofTnipaneabisetosa ME and T. nigricornis at the polymorphic loci, and EST. Allele-' VOLUME NUMBER 109, 2 301 (-) Rf 0.9 0.8 T. bisetosa T. nigricornis 00 100 Genotype (%) Fig. 3. Zymogram ofacid phosphatase (ACPH) in Triipanea nigricornis and T. bisetosa populations, showing a fixed allele difference between the two species. PGM lations than in T. bisetosa populations. This greater genetic variation for Four alleles for that locus were detected seemed to be preserved in T. nigricornis in T. bisetosa populations; six alleles populations since chi-square tests on were detected in T. nigricornis popula- expected and observed genotypic frequen- PGM tions. The mean number of alleles was cies indicated that the locus did not 2.67 in T. bisetosa samples and 5.00 in T. depart from the Hardy-Weinberg equilib- nigricornis samples. A summary of all rium (Table 4). The Hardy-Weinberg test PGM phenotypes and their genetic in- was not applied to T. bisetosa because the terpretation is presented in Fig. 4. The locus was almost fixed for the most two species shared the alleles 0.09, 0.12, common allele, leaving the other alleles 0.16, and 0.19. Allele 0.12 was the most with genotypic classes having low fre- common allele in both species, with quencies (<1), even after pooling. a total frequency of 0.964 in T. bisetosa The polymorphic locus, PGI, behaved and 0.609 in T. nigricornis (Table 3). as a dimer. It also showed genetic Two rare alleles (0.06 and 0.10) were variability between the species. Two only detected in T. nigricornis samples alleles were detected in T. bisetosa; five from flies reared from E.farinosa flower- were detected in T. nigricornis. A mean heads, but not E. fructescens heads. of 1.33 alleles per locus was found in T. Moreover, PGM was almost fixed in bisetosa samples while a mean of 3.33 three T. bisetosa samples with the fre- alleles per locus was found in T. nigri- quency of the most common allele (0.12) cornis samples. The phenotypes observed greater than 0.95 (Table 3). for PGI in the two species are summa- PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON 302 Table 3. Allele frequencies of Trupanea bisetosa and T. tiigricornis at two polymorphic loci, PGI and PGM. VOLUME 109, NUMBER 2 303 < m Q O ffl < o G o o in CQ On o CQ u u . PROCEEDINGS OF THE ENTOMOLOGICAL SOCIETY OF WASHINGTON 304 Table4. TestforHardy-Weinbergequilibriumat the PGM locusin T. nigricornispopulations. Alleles were pooled to maintain expected genotypic frequency > 1