The yellowjackets and hornets are prominent insects in temperate regions. The use of the term yellowjacket can have different regional interpretations. I use "yellowjacket" for the genera Vespula and Dolichovespula, and "hornet" for the genus Vespa, following Greene and Caron (1980).
Yellowjackets and hornets are in the subfamily Vespinae of the hymenopteran family Vespidae. Vespines are all eusocial, or derived from eusocial ancestors. Eusociality is defined by a reproductive division of labor (separate queen and worker castes), overlap of generations, and cooperative care of offspring (Wilson, 1971).
Colony cycle (Akre et al., 1980)
The colony cycle of non-parasitic yellowjackets and hornets of the temperate zone starts in Spring. A single queen, who has overwintered, initiates a new colony (this a contrast to some other social vespids, such as paper wasps, where a number of queens initiate a colony). The colony grows during the summer with the production of many non-reproductive female workers. All of the eggs are laid by the queen. In the fall the colony starts producing reproductive wasps. These are males and new queens. They are larger than workers in most species. The new queens mate with males and overwinter in a protected location, such as under the bark of a decaying log. The workers and the males of the colony die as winter begins.
Nests (Akre et al., 1980)
The nests of vespines can be located in many different places. They can be exposed above the ground, above ground in a crevice (e.g. a tree trunk), or subterranean, where the wasps excavate the soil. The location of the nest often varies within a species. There is a large range in colony size between species. Some species tend to be very succesful, and have nests that contain several thousand adults. Other species have much smaller nests. Vespine nests are usually surrounded by a masticated paper envelope. The envelope surrounds tiers of round combs. When tiers are added as the colony grows, the outer envelope is expanded. The architecture of nests has been used in phylogenetic studies (see vespine systematics).
Foraging (Akre et al., 1980)
The workers of vespine colonies do all of the foraging for food. Arthropod prey is used as a protein source which is fed to larvae. Some species only take live prey, whereas other species (especially those of the Vespula vulgaris group) will also scavenge for dead organisms. Vespines also feed on nectar at flowers as a carbohydrate source. Therefore, it is common to collect vespines at flowers.
There are several vespine species in North America that exhibit social parasitism. Obligate social parasites cannot initiate their own colonies, but must invade established nests of other species. The queen of the social parasite will kill the queen of the host colony and use the host colony's workers to produce her own offspring. There are some species which are facultative social parasites. The females of these species can start their own nests in the spring or they can be parasitic on another species. It is thought that facultative parasitism is an intermediate step in the evolution of obligate parasitism. In the Pacific Northwest, V. sulphurea is a facultative social parasite; Dolichovespula arctica and Vespula austriaca are obligate social parasites.
Systematics and Classification of the Vespoidea and Vespidae
The definition of the superfamily Vespoidea has changed over time. In earlier usages, Vespoidea referred to the group of closely related wasp lineages which are now included in the family Vespidae (Richards, 1962). Vespoidea currently refers to a group of families comprising a large section of the Aculeata. The main families in Vespoidea are Vespidae, Formicidae, Mutilidae, and Pompilidae (Brothers and Carpenter, 1993).
The classification of the family Vespidae has been unstable. What some classifications call subfamilies or tribes, others call families. Bradley (1922) and Bequaert (1928) had one family, the Vespidae, containing eleven subfamilies. Richards (1962) split Vespidae into three families, Masaridae, Eumenidae, and Vespidae in the superfamily Vespoidea. The revisionary work of Carpenter (1982) used cladistic phylogenetics to classify the vespoids. This revealed that Richards' families Masaridae and Eumenidae were paraphyletic. Carpenter (1982) proposed the lumping of the vespoid families of Richards into the single family Vespidae. The Vespidae of Carpenter (1982) contains six subfamilies: Euparagiinae, Masarinae, Eumeninae, Stenogastrinae, Vespinae, and Polistinae. Many of the subfamilies of Richards were synonymized.
The cladistic analysis of Vespoidea by Carpenter (1982) is summarized in Fig. 1. The Vespinae are the yellowjackets and hornets, treated here. The Polistinae are the paper wasps. In addition to morphological synapomorphies, the clade composed of the Polistinae and Vespinae is united by eusociality. The Stenogastrinae is mostly subsocial, but some of its members have developed eusociality independent of the Vespinae + Polistinae clade. The Stenogastrinae are only found in Southeast Asia. The remaining subfamilies (Euparagiinae, Masarinae and Eumeninae) have not evolved eusociality. The Eumeninae are the mason and potter wasps. They are the most speciose vespid group in North America, with 260 species (Borror et al. 1989). The clade (Eumeninae (Stenogastrinae (Polistinae + Vespinae))) is characterized by a crease in the wings, such that they can be folded longitudinally at rest. The Masarinae are solitary wasps that make mud nests on rocks or twigs (Borror et al. 1989). In North America they are found only in the western US and adjacent Canada (Akre et al. 1980). Their center of diversity is in the southern hemisphere (Brothers and Finnamore, 1993). The Masarinae are unique among the Vespidae in that they provision their cells with a mixture of pollen and nectar rather than arthropod prey. Euparagiinae is a rare nearctic group which is the sister group to all other vespids (Carpenter, 1982). Species of the genus Euparagia, the single genus in Euparagiinae, have retained many primitive characters. Some Euparagiinae make ground nests which they provision with beetle larvae (Brothers and Finnamore, 1993). They are found only in the western United States (Borror et al., 1989).
Classification of Vespinae
There is little question of the monophyly of the Vespinae. Carpenter (1982) showed eight synapomorphies defining the vespines. Easily seen morphological characters include the loss of the jugal lobe of the hind wing and the first metasomal tergum being truncate anteriorly (Brothers and Finnamore, 1993).
The current classification of the genera within the subfamily Vespinae varies. This report follows the classification of Carpenter (1987b), which recognizes the genera Vespa, Dolichovespula, Provespa and Vespula. The genera Vespa, Dolichovespula, and Vespula occur in the North America. The genus Provespa is found in the tropical regions of the Orient (Spradbery, 1973).
Genus Vespa Linnaeus, 1758
The genus Vespa is native to the old world. The members of this genus are considered the true hornets. V. crabro, the European hornet, has been introduced into North America.
Genus Dolichovespula Rohwer, 1916
The genus Dolichovespula can be differentiated from the genus Vespula by the length of the malar region, the distance between the bottom of the compound eye and the base of the mandibles (see couplet 2 of the technical key). There are four species of the genus Dolichovespula in the Northwest US. This includes one species, D. arctica, which is a social parasite. Wasps of this genus tend to have aerial, exposed, nests. However, they occasionally have nests in protected places, such as inside a tree trunk or underground. Dolichovespula maculata, the bald faced hornet, is a common species which has a large exposed nest.
Genus Vespula Thomson, 1869
The Vespula species of North America are contained in two species groups: those closely related to V. rufa and those related to V. vulgaris. There are some authors (e.g. Varvio-Aho et al., 1984) who recognize the two species groups of Vespula as genera, but I retain Carpenter's classification.
V. vulgaris species group
There are three species in the V. vulgaris species group in the Northwest US. The members of the V. vulgaris species group are some of the most abundant and pestiferous wasps. The wasps of this species group have evolved to be scavengers as well as predators. The success of these wasps has been attributed to these scavenging tendencies. Scavenging increases food possibilities in the fall, when live insect prey becomes scarce. This extends the colony life cycle later into the year, resulting in larger colonies. Colonies of this group can be active for two months longer in the fall than colonies of the V. rufa group and Dolichovespula (Akre et al., 1980).
V. rufa species group
There are five species in the V. rufa group in the Northwest. This includes one obligate social parasite (V. austriaca), and one facultative parasite (V. sulphurea). Wasps in this species group tend to be predacious only on live prey. The species in the V. rufa group tend to be less abundant and less pestiferous than the wasps of the V. vulgaris group.
Although I follow Carpenter, the placement of some species in particular species groups has been controversial, and alternative placements have been proposed (MacDonald and Matthews (1975, 1984)).
Systematics of the Vespinae
The understanding of the evolutionary history of the Vespinae has changed greatly as new techniques and trends have developed. Many studies have only looked at behavioral characters (Greene, 1979) or a combination of behavioral and morphological characters in estimating phylogeny Carpenter (1987b). The evolutionary history of the Vespinae has also been examined using techniques that do not rely on morphological or behavioral characters: protein electrophoresis (Varvio-Aho et al., 1984) and DNA sequencing (Schmitz and Moritz, 1990).
Behavioral characters have been used in vespine phylogenetic analyses by Carpenter (1987b), Matsuura and Yamane (1984) and Greene (1979). As in morphological characters, which are traditionally used, these behavioral characters can be scored for each taxon and then polarized, as ancestral or derived, using an outgroup. In vespines these characters include nest architecture and the behavior of the colony and the individual.
Greene (1979) formed a hypothesis of yellowjacket (the genera Vespula and Dolichovespula) phylogeny based exclusively on behavioral characters. In this study, Greene used two kinds of behavioral characters. One type was based on social structure. The work on insect societies by Wilson (1971) helps define the social character states which Greene (1979) considered advanced. A character that Greene (1979) considered important is the differentiation of workers from queens. In Dolichovespula there is no morphological difference between the castes; in Vespula there is differentiation. Another character that Greene used to show that Dolichovespula was "less sophisticated" than Vespula is the means of control of the worker caste. The Dolichovespula queen maintains dominance over the workers in the colony through physical means, eating eggs that workers may lay and defending cells so the workers will not lay eggs. In Vespula the queen maintains control over the workers through pheromones; this is considered more advanced.
The other type of behavioral character used is nest architecture. There are a number of different nest characters that have been used. They include the color and texture of the paper covering the nest and the way the tiers of comb are supported. Unlike characters of social development, these characters do not have obvious advanced and primitive states. Comparison with known primitive groups (out groups) is necessary.
Greene (1979) concluded that Dolichovespula is more closely related to Vespa than to Vespula. Yet he also stated that Vespula arose from Dolichovespula which arose from Vespa, which implies that Dolichovespula is more closely related to Vespula. This non-cladistic reasoning was noticed by Carpenter (1987a). Greene did not define groups by shared derived characters, as in cladistics, but instead based evolutionary history and relatedness of species soley on the degree of social sophistication.
Carpenter (1987b) used cladistics to analyze the relationships of the genera of Vespinae (Vespula, Dolichovespula, Provespa and Vespa), and the major species groups of Vespula (Vespula (S. Str.) (=V. rufa group), Paravespula (=V. vulgaris group), Rugovespula, and the V. squamosa group). In this analysis, seventeen morphological and seven behavioral characters were used. The polarity of some of the characters used in earlier phylogenies (e.g. Matsuura and Yamane, 1984; Greene, 1979) were reevaluated for use in this study. In the phylogeny proposed by Matsuura and Yamane (1984) Vespa was shown to be the sister group of a polytomy consisting of Dolichovespula, Vespula and Provespa. By reevaluating the characters of Matsuura and Yamane (1984), Carpenter (1987b) was able to resolve this polytomy.
Much of the behavioral character set of Carpenter (1987b) was based on the work of Greene (1979). Greene (1979) compiled behavioral data for the genera Vespula and Dolichovespula. Carpenter (1987b) applied this information in a cladistic analysis.
In the study of Carpenter (1987b) the behavioral and morphological data sets were analyzed separately as well as together. The morphological characters alone resulted in the phylogenetic tree shown in Fig. 2. Within the genus Vespula, the subgenera Rugovespula and Paravespula (=V. vulgaris group) form a clade. The sister group to this is a group containing Vespula (S. Str.) (=V. rufa group) and the V. squamosa species group. Dolichovespula is a sister group to Vespula (S. Lat.); Provespa is the sister group to this clade, and Vespa is the sister group to all the other vespines.
Carpenter (1987b) also analyzed the data set including both the morphological and behavioral characters. This gave the same tree as the morphological characters alone (Fig. 2), but it had a smaller consistency index. The analysis of the behavioral characters alone gave a completely unresolved consensus tree. Because the morphological characters alone resulted in a more supported tree than the combination of morphological and behavioral characters, Carpenter showed that the morphological characters explained the behavioral characters better than the converse.
Varvio-Aho et al. (1984) used electrophoretic analysis to estimate the phylogeny of eight palearctic vespines representing Dolichovespula and the V. rufa and V. vulgaris species groups of Vespula. This study found that the Vespula species groups, the vulgaris species group and the rufa species group, may not form a monophyletic group. This is first time that such a situation has been proposed. It goes against all the prior literature, including Carpenter (1978b) and Bequaert (1931). Carpenter (1987a) criticized the methods used in the study of Varvio-Aho et al. (1984). He reanalyzed their results and found all of their conclusions to be ambiguous.
Schmitz and Moritz (1990) used mitochondrial DNA to estimate the phylogeny of European Vespinae. This study supported the monophyly of the genera Vespula and Dolichovespula, but it did not support the monophyly of Vespula and Dolichovespula. This DNA analysis found Dolichovespula to be most closely related to Vespa. This differs from Carpenter (1987b) which shows Vespa as the ancestral group, and Dolichovespula and Vespula as most closely related. Within the genus Vespula Schmitz and Moritz (1990) found that V. germanica was more closely affiliated with the V. rufa species group than the V. vulgaris species group, which is the conventional grouping.
Carpenter (1992) criticized the method of analysis used by Schmitz and Moritz (1990). Schmitz and Moritz (1990) presented genetic distance data based on DNA sequences. The analysis of Schmitz and Moritz (1990) failed to meet the assumptions of the analytical method they were using. The method that they used, UPGMA, is no longer considered appropriate for estimating phylogeny (Quicke, 1993). The reanalysis by Carpenter (1992) showed that none of the conclusions of Schmitz and Moritz (1990) are supportable.
Page author: Matthew P. Kweskin firstname.lastname@example.org
Last modified: 2 Feb 1997