Postmating reproductive isolation between Chrysopa quadripunctata and Chrysopa slossonae: mechanisms and geographic variation.
Received May 23, 1995. Accepted October 31, 1995
Although the importance of reproductive isolation in the formation and maintenance of species has been widely accepted (Mayr 1942; Dobzhansky 1951; Littlejohn 1981; Futuyma 1986; Endler 1989; Ridley 1993; but see Paterson 1985), elucidating the genetics and transformation of specific reproductive isolating mechanisms remains a formidable challenge for evolutionary biologists (Coyne and Orr 1989a,b; Endler 1989; Coyne 1993). Not only do reproductive barriers frequently entail interactions between two or more genetically differentiated entities (e.g., incipient species or sister species), but each of these entities may evolve in response to an array of selection pressures, including those that emanate from the other. Furthermore, reproductive isolation encompasses a range of ecological, behavioral, and physiological traits that may serve multiple functions in the lives of the interacting organisms. Consequently, it is often very difficult to estimate either the degree to which individual traits act as reproductive isolating mechanisms or the extent to which there is direct or indirect selection on the traits for a reproductive isolating function (e.g., Moore 1949; Butlin 1989; Howard 1993).
Two types of investigations can help elucidate how reproductive isolation evolves: first, identification of traits that act as reproductive barriers to hybridization, and second, determination of how these traits vary among populations of the species under study. Previously, we demonstrated geographic variation in the ability of sister species of green lacewings (Insecta: Neuroptera: Chrysopidae) to interbreed in the laboratory (Tauber et al. 1993). One of these species is a generalist predator (Chrysopa quadripunctata Burmeister), and the other is a specialist (C. slossonae Banks). Although the specialist is fully sympatric with the generalist, no hybrids have been reported from nature. Given our results, we reasoned that barriers to interbreeding may not have evolved strictly as a secondary consequence of the specialist's adaptation to its specific prey (Tauber et al. 1993). Rather, we suggested that natural selection for prezygotic barriers to hybridization may have had a direct role in the evolution of the species' reproductive isolation.
To begin testing this hypothesis, our present study focuses on identifying the mechanisms that hinder interbreeding when individuals from the two sister species come into contact. Specifically we tested for behavioral and gametic barriers to reproduction, as well as reduced viability and fertility of hybrids. Also, we expanded the earlier investigation on geographic variation in reproductive isolation by adding a fourth C. quadripunctata population to the comparison. A separate study considers seasonal isolation between the species (Albuquerque et al., unpubl.).
MATERIALS AND METHODS
Experimental Animals and Rearing Procedures
Chrysopa quad ripunctata occurs throughout the United States and southern Canada (e.g., Banks 1903; Bickley and MacLeod 1956; Agnew et al. 1981; Tauber and Tauber 1987). Over its extensive range, this generalist predator feeds on a diverse array of prey that, in turn, feed on a variety of plants (Smith 1922; Agnew et al. 1981); a suite of life-history, behavioral and larval morphological traits differentiates it from C. slossonae (Milbrath et al. 1993, 1994; Tauber et al. 1995a,b; Albuquerque et al., unpubl.).
Because of the remarkable similarity between the adult morphology of the sister species, C. slossonae's geographical distribution is not fully established. It is reported from northeastern and mid-Atlantic United States (Pergande 1912; Bickley and MacLeod 1956; Tauber and Tauber 1987), recently we examined specimens from southeastern United States (South Carolina, Georgia, and northern Florida). Its single prey, the woolly alder aphid [Prociphilus tesselatus (Fitch)], ranges from southeastern Canada through the eastern United States into northern Florida. There is no indication that it occurs in the far western United States or in the central or southern Great Plains; an isolated record exists from Utah (Smith and Parron 1978). In our tests we used C. slossonae from New York, C. quadripunctata from two populations that are sympatric with C. slossonae (New York and Florida), and two that are allopatric (California and Kansas). Voucher specimens are in the Cornell University Insect Collection (Lot 1205).
Laboratory stock of both species originated from three to six field-collected females per site, per year of collection (see Table 1). First-generation, laboratory-reared offspring from each female were allocated in approximately equal numbers across all types of pairings. Larvae were reared individually, and to avoid inbreeding, siblings were not paired. Each larva received a continuous supply of green peach aphids, Myzus persicae (Sulzer), pea aphids, Acyrthosiphon pisum (Harris), and eggs of the moth Sitotroga cerealella (Olivier). Adults were paired in cardboard cages (one pair per cage) and provided distilled water, a protein-carbohydrate mixture (1:1: 1:1 volumetric mixture of sugar, honey, protein hydrolysate of yeast, and Wheast[R]), and a continuous supply of prey. Prey for adults of both species consisted of pea aphids and green peach aphids. For C. slossonae adults, we also added their usual prey, woolly alder aphids. All experiments were carried out under constant conditions of photoperiod (L:D 16:8) and temperature (24 [+ or -] 1 [degrees]C). Because it was not possible to test all populations concurrently, experiments were conducted in the summer and fall of 1990-1993.
Species/population Locality (years of collection) C. quadripunctata New York (NY) Ithaca, Tompkins Co. (1990-1993) Florida (FL) Monticello, Jefferson Co. (1990) Kansas (KS) Tuttle Creek Dam, Pottawamie Co. (1991,1992) California (CA) Davis, Yolo Co. (1990-1992) C. slossanae New York (NY) Ithaca, Tompkins Co. (1990-1993) Species/population Plant association C. quadripunctata New York (NY) Quercus sp., Carya sp., A. incana Florida (FL) Carya illinoessis (Wang.) K. Koch Kansas (KS) Quercus macrocarpa Michx. California Quercus lobata Nee C. slossanae New York (NY) Alnus incana ssp. rugosa (Du Roi) Clausen
Experiment I: Geographic Variation in the Ability of C. quadripunctata to Hybridize with C. slossonae.--This experiment had two parts. First, we measured the ability of females and males from each of the four geographic populations of C. quadripunctata (New York, Florida, Kansas, and California) to interbreed with C. slossonae from New York, as well as with conspecifics from the other populations, in no-choice situations. Newly emerged first-generation, laboratory-reared females were paired with a single interspecific male or a conspecific male from either the same or a different population, 10-30 pairs per type of pairing. For every cross, we had appropriate concurrent intrapopulation or intraspecific pairs as controls. Because parthenogenesis is unknown in Neuroptera (Chapman 1982), production of fertile eggs, i.e. eggs in which embryos developed, served as the criterion for successful reproduction (i.e. the occurrence of mating, transfer of sperm, fertilization, and oviposition). Infertile females were those that did not oviposit or that laid eggs that did not develop embryos; no mechanisms are implied in the term. Each pair was observed daily for oviposition; after oviposition began, egg fertility was determined from a sample of 10-15 eggs per pair.
In the second part of Experiment I, we examined the viability (stage-specific survival from oviposition to adult emergence), sex ratio and fertility of the interspecific and interpopulation hybrids. Survival and sex ratios were assessed from a sample of 10-15 offspring per pair.
The fertility of hybrids was examined by pairing the [F.sup.1] males and females that originated from the interspecific crosses (except those involving C. quadripunctata from Kansas). When infertility in [F.sup.1] hybrids is sexually asymmetrical, it is usually higher in the heterogametic sex (Haldane 1922). Therefore, we tested the ability of hybrid (C. quadripunctata x C. slossonae) males [the heterogametic sex in Chrysopidae (Semeria 1984)] to backcross with pure C. quadripunctata females. All tests included concurrent controls ([F.sub.1], intrapopulation pairs). Five pairs per cross were tested; successful reproduction was determined from a sample of 25 eggs per pair.
Experiment II: Mechanisms Underlying Reproductive Isolation.--Here we investigated the mechanisms that reduce interbreeding when C. quadripunctata and C. slossonae are paired in the laboratory. It focused on C. quadripunctata females and C. slossonae males from New York because of the low incidence of hybridization between individuals from these populations. We also tested approximately one third of the nonovipositing females from crosses involving the other geographic populations.
Experiment II had two parts. First, to determine whether the barrier to hybridization occurs before or after mating, we paired twenty C. quadripunctata females with twenty C. slossonae males, one pair per cage. After 10 days, we dissected half of the females, chosen at random, and examined them for the presence of sperm; the remaining females were held for an additional 10 days before dissection. Any eggs that were laid were tested for fertility. For comparison, females and eggs from intraspecific pairs of both species were handled in the same manner.
Dissections were made in physiological saline. After observing the developmental stages of the oocytes, we removed the internal reproductive organs (except the ovaries) and viewed them under high power (X400) for the presence of sperm. Sperm was visible through the walls of the internal organs (the bursa copulatrix and spermatheca).
The second part of Experiment II investigated the paternity of offspring when interspecific males, in crosses with C. quadripunctata females, were replaced by conspecific males. This test was possible because interspecific hybrid larvae differ in head and body markings from the larvae of either parental type (Tauber et al. 1995a).
Twenty interspecific pairs (C. quadripunctata females and C. slossonae males) were provided prey to promote egg development, mating, and oviposition. Females that did not produce fertile eggs within 10 days after pairing (a period sufficient for reproduction) received conspecific males in place of the C. slossonae males. To determine the paternity of subsequent offspring, the first 30 eggs from each female, as well as 20 eggs laid one week later, were reared and the head markings of the resulting larvae were examined. For comparison, we also reared 30 offspring per female from intraspecific (control) pairs.
Finally, we tested whether C. quadripunctata females that had mated with a C. slossonae male and produced hybrid offspring would subsequently mate with a conspecific male and produce pure C. quadripunctata offspring. Chrysopa slossonae males were removed from the four interspecific pairs in which the C. quadripunctata females had produced fertile eggs, and each was replaced with a conspecific male (about 25 days after the first oviposition). Then, we determined the paternity of subsequent offspring (hybrid versus pure C. quadripunctata) by rearing 20 eggs per female per day for one week after male substitution, and then 20 eggs per week every week until the females died.
We used the G-test of independence with Yates correction to compare the frequencies of hybridization (Sokal and Rohlf 1981); the proportion from each interpopulation or interspecific cross was compared with that from its reciprocal cross, as well as with those from the two respective intrapopulation (control) pairings. To test for departures from the expected sex ratio of 1:1 in the offspring from each of the crosses, we used the G-test for goodness of fit (Sokal and Rohlf 1981).
We analyzed the data on stage-specific survival with one-way ANOVA; frequencies were arcsine transformed and means were separated with the Tukey method of multiple comparisons, GLM Procedure (SAS Institute 1985). The experimental unit was the pair; i.e., we averaged the multiple measurements performed on the offspring from each pair. The level of significance in all tests was 0.05.
Experiment I: Geographic Variation in C. quadripunctata's Ability to Hybridize with C. slossonae
Hybridization Rates.--Two types of crosses showed significantly lower incidences of hybridization than either control, i.e. when C. quadripunctata females from New York and Florida were paired with C. slossonae males (Fig. 1). Although there were significant differences in hybridization among the other crosses, none deviated significantly from the C. slossonae control. Intraspecific pairings among individuals from the four populations of C. quadripunctata had incidences of fertile oviposition as high as the intrapopulation pairings (controls), and there were no differences between males and females in the ability to hybridize with individuals from any of the four populations. The results provide no suggestion that infertility was particularly high among the offspring of specific pairs.
[Figure 1 ILLUSTRATION OMITTED]
Viability and Sex Ratio of Hybrids.--Despite the variation in fertility among interspecific pairs, hybrid immatures from all crosses had survival rates (egg to adult) that were similar to those of pure C. quadripunctata (including all inter- and intrapopulation combinations) and pure C. slossonae (Fig. 2). The sex ratio of offspring from most interspecific and intraspecific pairs did not diverge significantly from 0.5 (range = 0.38-0.62); the only exception occurred in the test involving C. quadripunctata females from Kansas and males from New York, which yielded 61% females (G-test, I df, G = 4.856, P [is less than] 0.05).
[Figure 2 ILLUSTRATION OMITTED]
Hybrid Fertility.--Most [F.sub.1] intrapopulation pairs produced fertile eggs, as did [F.sub.1] hybrids stemming from C. slossonae and C. quadripunctata from California (Table 2). Incidences of fertile oviposition in [F.sub.1] hybrids involving C. quadripunctata from New York or Florida ranged from 0 to 40%.
[TABULAR DATA 2 NOT REPRODUCIBLE IN ASCII]
Experiment II: Mechanisms Underlying Reproductive Isolation Hybridization Barrier: Pre- or Postmating?
When C. quadripunctata females were paired with C. slossonae males (both populations from New York), only a low proportion (10%) of the females produced fertile eggs within 10 days of pairing (Table 3). Similarly, only 22% of the females were fertile after twenty days of pairing. Upon dissection, all females had mature oocytes in the ovarioles and mobile sperm in the bursa copulatrix, but only the females that had laid fertile eggs had sperm in their spermathecae (one female from the 10-day group had sperm in the spermatheca but did not lay fertile eggs). A great majority of the females from intraspecific pairs of both age groups laid fertile eggs; all had sperm in the spermatheca.
[TABULAR DATA 3 NOT REPRODUCIBLE IN ASCII]
Dissections of one-third of the infertile C. quadripunctata females from populations other than New York (Experiment I) showed a similar postmating barrier to hybridization. All non-ovipositing females had sperm in their bursae copulatrices but not in their spermathecae. These C. quadripunctata females included: (a) one from New York paired with a conspecific male from California; and (b) four from Kansas and one from California crossed with C. slossonae males (Fig. 1). Similarly, C. slossonae females that did not lay fertile eggs when paired with C. quadripunctata males, had sperm only in the bursa copulatrix, i.e. three females paired with New York males, one with a Kansas male, and three with California males.
Male Replacement and Offspring Type.--All C. quadripunctata females that were infertile in the first part of Experiment II (Table 4) laid fertile eggs within one day of being paired with a conspecific male. All of the resulting offspring were pure C. quadripunctata. Given the results from our dissections (Table 3), it is very likely that these females had heterospecific sperm in their bursae copulatrices when they were subsequently paired with conspecific males.
[TABULAR DATA 4 NOT REPRODUCIBLE IN ASCII]
Among the four C. quadripunctata females that had hybridized with C. slossonae in the first part of Experiment II, replacement of the heterospecific male with a conspecific one had two results (Table 4). Two of the four females continued to yield hybrid offspring for about 15 days after male replacement and then began producing pure C. quadripunctata offspring. The other two females laid hybrid eggs for about 30 days, after which they died; there was no production of conspecific offspring.
Because of habitat and phenological differences, sexually active C. quadripunctata and C. slossonae may seldom encounter each other in nature (Tauber and Tauber 1987; Albuquerque et al., unpubl.). Nevertheless, our current results show that when the two species are brought together in the laboratory, neither behavioral nor mechanical barriers prevent mating or the transfer of sperm to females. Indeed, the incidence of sperm transfer in interspecific pairs equalled that of conspecific pairs (Table 3). However, a postmating, prezygotic mechanism (gametic barrier) reduced the level of hybridization between the two species--especially between C. quadripunctata females and C. slossonae males.
Natural hybrids have not been found in the field; but when hybrids resulted from crosses in the laboratory, they were viable and their sex ratios did not differ from 1:1 (Fig. 2). Moreover, hybrids whose parents were derived from sympatric populations of C. quadripunctata and C. slossonae showed low fertility ([is less than or equal to] 40%; Table 2). Thus, hybrid sterility may constitute another barrier to gene exchange between the two species; hybrid breakdown also remains a possibility.
Postmating, Prezygotic Barrier to Hybridization
Among insects with internal insemination, the male generally transfers sperm to the female's bursa copulatrix during copulation. Subsequently, the sperm move or are moved to the storage organ, the spermatheca, from which they are released prior to egg fertilization (Chapman 1982). Secretions from male accessory glands may be transferred with the sperm and may promote reproduction (Chapman 1982, Engelmann 1984; Davey 1985; Principi 1985). Chrysopa falls readily into this pattern. All C. quadripunctata and C. slossonae females from both fertile conspecific and fertile interspecific pairings contained sperm in their spermathecae (Table 3). Infertility was associated with sperm retention in the bursa copulatrix, very little oviposition, and the absence of fertile eggs; this situation typifies gametic isolation (see Mayr 1963; Dobzhansky 1970).
The gametic barrier occurred in some females from each of the geographical populations of C. quadripunctata that we tested and also in C. slossonae females, but it affected the incidence of fertile reproduction significantly only when C. slossonae males were paired with C. quadripunctata females that originated from areas where the species occur sympatrically (Fig. 1). Because all of the infertile females that we dissected had sperm in their bursae copulatrices, but not in their spermathecae, we concluded that it was the failure to transfer sperm into the spermatheca that resulted in the lack of fertile oviposition.
In the past, studies of reproductive isolation at the gametic level have largely focused on Drosophila (e.g. Patterson 1946, 1947; see also Dobzhansky 1970 for a review). More recently, investigations with other groups of insects, such as Orthoptera, Coleoptera, and now Neuroptera (e.g. Katakura 1986; Juberthie-Jupeau 1988; Hewitt et al. 1989; Bella et al. 1992; Howard and Gregory 1993; Wade et al. 1994) have begun to reveal the diverse mechanisms that this category encompasses. In most cases the specific nature of the reproductive barrier is not well understood.
The mechanisms that have been proposed for gametic isolation cover the full spectrum of steps between copulation and egg fertilization. They range from lack of sperm transfer during copulation (e.g., Grimaldi et al. 1992) to incompatibility between the sperm and the cytoplasm of the egg (e.g. Laven 1967). But most of the mechanisms investigated to date involve a negative effect on the sperm, such as the suppression of sperm mobility (Patterson and Stone 1952; Koref-Santibanez 1964; Gregory and Howard 1994) or sperm inactivation and expulsion (Dobzhansky et al. 1968). In some cases, enlargement and hardening of the vagina obstructs sperm movement into the spermatheca (Patterson 1946 1947).
Although the gametic barrier to hybridization between C. quadripunctata and C. slossonae may involve a negative effect on the sperm, it does not readily fall into the categories of mechanisms described thus far. First, there are no visible changes in the bursa copulatrix after interspecific mating and transfer of heterospecific sperm; and second, the sperm remain alive and mobile in the female for a relatively long period (at least 20 days). Therefore, we suggest that the gametic barrier in Chrysopa may occur when the female fails to perceive a species-specific chemical or other signal that influences the break-down of the spermatophore (see Principi 1985) or sperm transfer to the spermatheca.
Such a signal could act as a general stimulus (see Gupta and Smith 1969; Chapman 1982; Garbers 1989). But assuming that sperm removal does not occur in the C. quadripunctata-C. slossonae system, the production of uniformly conspecific offspring by previously infertile females in the male-replacement experiment (Table 4) is consistent with a species-specific chemical interaction between the sperm and the female. Such an interaction is one of several mechanisms that may explain the reproductive isolation between subspecies of grasshoppers (Hewitt et al. 1989; Bella et al. 1992). It also should be considered for other cases, e.g. cave beetles (Juberthie-Jupeau 1988) and some species of Drosophila in which sperm may remain in the bursa copulatrix for extended periods (see Nonidez 1920; Coyne 1993).
The results from our male-replacement experiment indicated that C. quadripunctata females that had previously hybridized with C. slossonae may remate with a conspecific male and switch from the production of hybrids to the production of conspecific offspring. However, the switch occurred after a considerable delay (two weeks), and not all females did so (Table 4). The mechanisms involved and the implications of these findings to reproductive isolation remain to be investigated. They are consistent with a variety of explanations, such as sperm depletion, sperm displacement, sperm competition, or sperm precedence (see e.g. Thornhill and Alcock 1983).
Hybrid Viability and Fertility
Our laboratory experiments gave no evidence that [F.sub.1] have reduced survival or abnormal sex ratios (Fig. 2). In the crosses that we tested (n = 5/cross), fertility was low ([is less than or equal to] 40%) in hybrids whose C. quadripunctata parent originated from a population that overlaps geographically with C. slossonae (New York and Florida; Table 2). In contrast, fertility was high (100%) in hybrids involving C. quadripunctata parents from a population that is allopatric to C. slossonae (California). Because all of these hybrid pairings were done before we discovered the barrier to sperm transfer (see also Tauber and Tauber 1987), we did not dissect the infertile hybrid females; whether sperm was retained in their bursae copulatrices is unknown.
Evolution of the Barrier to Sperm Transfer
The occurrence of sperm retention in some interpopulation C. quadripunctata pairings (Fig. 1) indicates that the mechanism that controls movement of sperm between the bursa copulatrix and the spermatheca is general and phenotypically variable among C. quadripunctata individuals. Consequently there is the distinct possibility that C. quadripunctata harbors genetic variation for the gametic barrier to hybridization and that natural selection may act on this variation to increase the sensitivity of the mechanism in populations of C. quadripunctata that overlap with C. slossonae.
Three characteristics of the C. quadripunctata-C. slossonae system are consistent with a model of direct selection for traits that retard hybridization in sympatric populations (reinforcement or reproductive character displacement; see Fisher 1930; Dobzhansky 1940; Butlin 1989; Howard 1993). First, interspecific pairs from sympatric populations (New York and Florida) have relatively high levels of reproductive incompatibility, whereas those from allopatric populations (Kansas and California) show high levels of hybridization (see also Tauber et al. 1993). Second, although the barrier to interbreeding occurs after mating, it intervenes before hybrids are formed. Third, hybrids have traits that generally are intermediate between those of their parents (Tauber and Tauber 1987; Albuquerque et al., unpubl. data); as a result, they would probably have low survival and low competitive ability in the habitats of either species in the field.
Given the above, the pattern of geographic and sexual variation in the postmating barrier to sperm transfer indicates that C. quadripunctata females constitute the main target of selection for this trait; neither C. quadripunctata males nor either sex of C. slossonae are involved. Moreover, the barrier to hybridization occurs after C. slossonae males have transferred sperm to C. quadripunctata females. As such, it does not confer an adaptive advantage to the males in terms of conserving time, energy, or gametes.
The question then arises: how could natural selection favor the evolution of a postmating barrier to hybridization that is restricted to C. quadripunctata females? Chrysopa quadripunctata females occasionally occur in C. slossonae's habitat (alder trees infested with woolly alder aphids), where they may be exposed to C. slossonae males (Tauber and Tauber 1987; Milbrath et al. 1994). Given that C. slossonae males are larger than C. quadripunctata males (Albuquerque et al., unpubl.), they may have a competitive advantage in courting and mating with either C. quadripunctata or C. slossonae females. Thus, it is not unreasonable to expect occasional interspecific (C. quadripunctata female x C. slossonae male) matings; in such cases, the postmating barrier to hybridization may serve to prevent the C. quadripunctata females from wasting reproductive effort in the production of hybrid offspring.
We thank R. G. Harrison (Cornell University), J. P. Nyrop (Cornell University, NYSAES), and J. Seger (University of Utah) for their thoughtful comments on the manuscript; C. E. McCulloch and S. J. Schwager (Cornell University) for advice on the statistics; J. R. Nechols, R. F. Mizell III, L. E. Ehler, and M. V. Kimsey for their help and cooperation; and the U.S. National Museum of Natural History (O. S. Flint Jr.), the Florida State Collection of Arthropods (L. A. Stange), the Mississippi Entomological Museum (R. L. Brown), and the University of Georgia Insect Collection (J. V. McHugh) for loan of specimens. We acknowledge the support of the National Science Foundation (Grant BSR 88-17822), Regional Project 185, Hatch Project 408 (MJT and CAT), Fellowship 204347/89-0 from the Conselho Nacional de Desenvolvimento Cientifico e Tecnologico, Brazil (GSA), and the Grace H. Griswold Fund (Entomology, Cornell University).
AGNEW, C. W., W. L. STERLING, and D. A. DEAN. 1981. Notes on the Chrysopidae and Hemerobiidae of eastern Texas with keys for their identification. Southwest. Entomol., Suppl. 4:1-20
BANKS, N. 1903. A revision of the Nearctic Chrysopidae. Trans. Am. Entomol. Soc. 29:137-162.
BELLA, J. L., R. K. BUTLIN, C. FERRIS, and G. M. HEWITT. 1992. Asymmetrical homogamy and unequal sex ratio from reciprocal mating-order crosses between Chorthippus parallelus subspecies. Heredity 68:345-352.
BICKLEY, W. E., and E. G. MAcLEoD. 1956. A synopsis of the Nearctic Chrysopidae with a key to the genera. Proc. Entomol. Soc. Wash. 58:177-202.
BUTLIN, R. 1989. Reinforcement of premating isolation. Pp. 158179 in J. A. Endler and D. Otte, eds. Speciation and its consequences. Sinauer, Sunderland, MA.
CHAPMAN, R. E 1982. The insects: structure and function. 3d ed. Harvard Univ. Press, Cambridge, MA.
COYNE, J. A. 1993. The genetics of an isolating mechanism between two sibling species of Drosophila. Evolution 47:778-788.
COYNE, J. A. and H. A. Orr. 1989a. Patterns of speciation in Drosophila. Evolution 43:362-381.
--. 1989b. Two rules of speciation. Pp.180-207 in J. A. Endler and D. Otte, eds. Speciation and its consequences. Sinauer, Sunderland, MA.
DAVEY, K. G. 1985. The female reproductive tract. Pp. 15-36 in G. A. Kerkut and L. I. Gilbert, eds. Comprehensive insect physiology, biochemistry and pharmacology. Vol. 1. Pergamon, Oxford.
DOBZHANSKY, TH. 1940. Speciation as a stage in evolutionary divergence. Am. Nat. 74:312-321.
-- 1951. Genetics and the origin of species. 2d ed. Columbia Univ. Press, New York.
-- 1970. Genetics of the evolutionary process. Columbia Univ. Press, New York.
DOBZHANSKY, TH., L. EHRMAN, and P. A. KASTRITSIS. 1968. Ethological isolation between sympatric and allopatric species of the obscura group of Drosophila. Anim. Behav. 16:79-87.
ENDLER, J. A. 1989. Conceptual and other problems in speciation Pp. 625-648 in J. A. Endler and D. Otte, eds. Speciation and its consequences. Sinauer, Sunderland, MA.
ENGELMANN, E 1984. Reproduction in insects. Pp. 113 - 147 in C. B. Huffaker and R. L. Rabb, eds. Ecological entomology. John Wiley and Sons, New York.
FISHER, R. A. 1930. The genetical theory of natural selection. Clarendon Press, Oxford.
FUTUYMA, D. J. 1986. Evolutionary biology. 2d ed. Sinauer, Sunderland, MA.
GARBERS, D. L. 1989. Molecular basis of fertilization. Annul Rev. Biochem. 58:719-742.
GREGORY, R G., and D. J. HOWARD. 1994. A postinsemination barrier to fertilization isolates two closely related ground crickets. Evolution 48:705-710.
GRIMALDI, D., A. C. JAMES, and J. JAENIKE. 1992. Systematics and modes of reproductive isolation in the Holarctic Drosophila testacea species group (Diptera: Drosophilidae). Ann. Entomol. Soc. Am. 85:671-685.
GUPTA, B. L., and D. S. SMITH. 1969. Fine structural organization of the spermatheca in the cockroach, Periplaneta americana. Tiss. Cell 1:295-324.
HALDANE, J. B. S. 1922. Sex ratio and unisexual sterility in hybrid animals. J. Genet. 12:101-109.
HEWITT, G. M., R MASON, and R. A. NICHOLS. 1989. Sperm precedence and homogamy across a hybrid zone in the alpine grasshopper Podisma pedestris. Heredity 62:343-353.
HOWARD, D. J. 1993. Reinforcement: origin, dynamics, and fate of an evolutionary hypothesis. Pp. 46-69 in R. G. Harrison, ed. Hybrid zones and the evolutionary process. Oxford Univ. Press, New York.
HOWARD, D. J., and P. G. GREGORY. 1993. Post-insemination signalling systems and reinforcement. Phil. Trans. Roy. Soc. Lond. B 340:231-236.
JUBERTHIE-JUPEAU, L. 1988. Mating behaviour and barriers to hybridization in the cave beetle of the Speonomus delarouzeei complex (Coleoptera, Catopidae, Bathysciinae). Int. J. Speleol. 17:
KATAKURA, H. 1986. Evidence for the incapacitation of heterospecific sperm in the female genital tract in a pair of closely related ladybirds (Insecta, Coleoptera, Coccinellidae). Zool. Sci. 3:115-121.
KOREF-SANTIBANEZ, S. 1964. Reproductive isolation between the sibling species Drosophila pavani and Drosophila gaucha. Evolution 18:245-251.
LAVEN, H. 1967. Speciation and evolution in Culex pipiens. Pp. 251-275 in J. W. Wright and R. Pal, eds. Genetics of insect vectors of disease. Elsevier, Amsterdam, Netherlands.
LITTLEJOHN, M. J. 1981. Reproductive isolation: a critical review. Pp. 298-334 in W. R. Atchley and D. S. Woodruff, eds. Evolution and speciation: Essays in honor of M. J. D. White. Cambridge Univ. Press, Cambridge.
MAYR, E. 1942. Systematics and the origin of species from the viewpoint of a zoologist. Dover, New York.
-- 1963. Animal species and evolution. Harvard Univ. Press, Cambridge, MA.
MILBRATH, L. R., M. J. TAUBER, and C. A. TAUBER. 1993. Prey specificity in Chrysopa: An interspecific comparison of larval feeding and defensive behavior. Ecology 74:1384-1393.
-- 1994. Larval behavior of predacious sister-species: Orientation, molting site, and survival in Chrysopa. Behav. Ecol. Sociobiol. 35:85-90.
MOORE, J. A. 1949. Patterns of evolution in the genus Rana. Pp. 315-338 in G. L. Jepsen, E. Mayr, and G. G. Simpson, eds. Genetics, paleontology, and evolution. Princeton Univ. Press, Princeton, NJ.
NONIDEZ, J. E 1920. The internal phenomena of reproduction in Drosophila. Biol. Bull. 39:207-230.
PATERSON, H. E. H. 1985. The recognition concept of species. Pp. 21-29 in E. Vrba, ed. Species and speciation. Transvaal Mus. Monogr. No. 4, Pretoria.
PATTERSON, J. T. 1946. A new type of isolating mechanism in Drosophila. Proc. Nat. Acad. Sci. 32:202-208.
-- 1947. The insemination reaction and its bearing on the problem of speciation in the mulleri subgroup. Univ. Texas Publ. 4720:41-77.
PATTERSON, J. T., and W. S. STONE. 1952. Evolution in the genus Drosophila. Macmillan, New York.
PERGANDE, T. 1912. The life history of the alder blight aphis. Tech. Bull. USDA 24:1-28.
PRINCIPI, M. M. 1985. Lo spermatoforo nei Neurotteri Crisopidi. Frustula Entomol. Nuova Ser., Vol. 7-8 (20-21):143-159.
RIDLEY, M. 1993. Evolution. Blackwell, Oxford.
SAS INSTITUTE. 1985. SAS user's guide: Basics. Vers. 5. SAS Institute, Inc., Cary, NC.
SEMERIA, Y. 1984. Some caryotypes in Chrysopidae. Pp. 42-48 in M. Canard, Y. Semeria, and T. R. New, eds. Biology of Chrysopidae. Dr. W. Junk Publishers, The Hague, The Netherlands.
SMITH, C. E, and C. S. PARRON. 1978. An annotated list of Aphididae (Homoptera) of North America. North Carolina Agricultural Experiment Station, Technical Bulletin 255, Raleigh.
SMITH, R. C. 1922. The biology of the Chrysopidae. Cornell Univ. Ag. Exp. Sta., Mem. 58:1287-1372.
SOKAL, R. R., and E J. ROHLF. 1981. Biometry. 2d ed. Freeman, New York.
TAUBER, C. A., and M. J. TAUBER. 1987. Food specificity in predacious insects: A comparative ecophysiological and genetic study. Evol. Ecol. 1:175-186.
TAUBER, M. J., C. A. TAUBER, J. R. RUBERSON, L. R. MILBRATH, and G. S. Albuquerque. 1993. Evolution of prey specificity via three steps. Experientia 49:1113-1117.
TAUBER, C. A., J. R. RUBERSON, and M. J. TAUBER. 1995a. Size and morphological differences among the larvae of two predacious species and their hybrids. Ann. Entomol. Soc. Am. 88: 502-511.
TAUBER, C. A., M. J. TAUBER, and L. R. MILBRATH. 1995b. Individual repeatability and geographical variation in the larval behaviour of the generalist predator, Chrysopa quadripunctata. Anim. Behav. 50:1391-1403.
THORNHILL, R., and J. ALCOCK. 1983. The evolution of insect mating systems. Harvard Univ. Press, Cambridge, M A.
WADE, M.J., H. PATTERSON, N. W. CHANG, and N. A. JOHNSON. 1994. Postcopulatory, prezygotic isolation in flour beetles. Heredity 72:163-167.
Gilbert S. Albuquerque, Department of Entomology, Comstrock Hall, Cornell University, Ithaca, New York 14853-0901; Present address: Laboratorio de Controle Biologico, Universidade Estadual do Norte Fluminense, 28015-620 Campos dos Goytacazes, Rio de Janeiro, Brazil.
Catherine A. Tauber, Department of Entomology, Comstrock Hall, Cornell University, Ithaca, New York 14853-0901
Maurice J. Tauber, Department of Entomology, Comstrock Hall, Cornell University, Ithaca, New York 14853-0901; E-mail: email@example.com