Arceuthobium has a base chromosome number of x = 14 (i.e., the lowest extant haploid number), and polyploidy is unknown (Wiens 1968). The haploid chromosome number (n = 14) can also be considered as a tetraploid number on a base of 7, but there is little evidence for this. The lowest known haploid chromosome number in the Viscaceae is n = 10, and it is best interpreted as the lower end of a decreasing aneuploid series from 14 to 10 in Viscum. The base number of x = 14 also characterizes the other New World genera of Viscaceae (Dendrophthora and Phoradendron), as well as Korthalsella, a widely distributed genus occurring in East Africa, southern Asia, and Oceania as far east as Hawaii. (Notothixos is karyologically distinct with n = 12; and the chromosomes of Ginalloa are unstudied). Chromosome evolution in the Viscaceae is discussed in detail by Wiens and Barlow (1971) and Barlow (1983, 1984). If polyploidy were involved in the evolution of x = 14 as a base chromosome number in Viscaceae, then its origin was certainly ancient. Under any circumstances, however, the chromosome systems in Arceuthobium and other genera of Viscaceae clearly operate as functional diploids.
The genetic system in Arceuthobium is therefore sexual, diploid, and obligately outcrossing because of dioecy. Vegetative reproduction and agamospermy are unknown (Player 1979), and there is virtually no manifestation of the occasional bisexuality that often characterizes dioecious groups. In fact, the structures of the non-functional sex are either absent or vestigial. The relatively high diploid chromosome number, concomitantly large number of linkage groups, and relatively high chiasma frequency (averaging perhaps 2 to 3 per bivalent at late diakinesis-early metaphase I) result in high levels of genetic variability.
High levels of genetic variation in Arceuthobium have been verified by electrophoretic studies of enzyme polymorphisms (Nickrent 1984, Linhart 1984, Nickrent and others 1984). The maintenance of high levels of genetic variability may be especially important among parasites. If the host population is also recombining genetically and producing novel genotypes, then the dwarf mistletoe must respond similarly to remain adapted. Such situations are known among scale insects, which become adapted to certain populations and even individuals of Pinus ponderosa (Edmunds and Alstad 1978).
Linhart (1984) presented preliminary evidence showing significant interpopulational electrophoretic variability in Arceuthobium vaginatum subsp. cryptopodum in Colorado; high variability has also been demonstrated for the host Pinus ponderosa (Mitton and others 1977) and for pines in general (Hamrick and others 1981). This pattern of interpopulational genetic variation suggests that dwarf mistletoes may also be adapting to local populations of its host, just as are the scale insects mentioned above.
In addition to maintaining population adaptiveness on genetically variable host populations, a diploid genetic system with high recombination potential should also be theoretically well suited for rapid evolution. A large number of ecological niches (host species) became available for colonization by dwarf mistletoes following the extensive speciation in North American pines (especially in Mexico) during the Miocene Epoch (Mirov 1967). The high proportion of North American Pinaceae presently parasitized by Arceuthobium (chapter 6) indicates that such an adaptive radiation has, in fact, occurred. With the exception of the eastern pines of North America, most of the widely distributed species of Pinaceae are parasitized by Arceuthobium (chapter 5).
Arceuthobium is primarily a parasite of Pinus—33 of the 47 taxa occur principally on this genus (chapter 6). Of these 33 taxa, 23 parasitize Diploxylon and 9 Haploxylon pines. Mirov (1967) recognizes 73 Diploxylon and 32 Haploxylon pines, so about one-third of each group is parasitized. The oldest known pine fossils are from the Jurassic Period, and even then the pines had differentiated into the subgenera Haploxylon and Diploxylon (Mirov 1967).
In both the Old and New Worlds, Pinus (Diploxylon and Haploxylon), Abies, and Picea are hosts of Arceuthobium. However, 3 other genera that occur in both New and Old Worlds are parasitized only in the New World (Larix, Pseudotsuga, and Tsuga). Juniperus, which is circumpolar, is a host for 3 of the 8 species of Arceuthobium found in the Old World; but junipers are not parasitized in the New World. The Old World genus Keteleeria is a host for dwarf mistletoe in southwestern China.
We originally suggested that the closest relative of Arceuthobium was Korthalsella (Hawksworth and Wiens 1972). Molecular data do not support such a relationship, but indicate that Notothixos is most closely allied to Arceuthobium (chapter 15). Resemblances between Korthalsella and Arceuthobium are no doubt the result of convergence. If the tropical Asian genus Notothixos and Arceuthobium had originated from a common ancestor, then Arceuthobium would have spread in the early Tertiary Period westward to the Mediterranean region (and possibly into North America), south into East Africa, and finally during the Miocene Epoch northeastward through Beringia into the New World (but see chapter 15). Thus Arceuthobium may have reached the New World in two migratory episodes—first in the early Tertiary Period by land connections with Europe (represented by A. abietis-religiosae and A. verticilliflorum as relict species) and second during the Miocene Epoch, which initiated the large radiation of species in Mexico and the western United States.
Generally, angiosperms in the Tertiary Period have evolved more rapidly than gymnosperms (Leopold 1967). Consistent with this trend, the dwarf mistletoes probably also evolved more rapidly than did their coniferous hosts.
Another factor in this adaptive radiation was the apparent lack of competition with other plants for these open niches (conifer hosts). With the possible exception of Phoradendron pauciflorum, which sometimes infects Abies concolor along with Arceuthobium abietinum, and a few tropical species of Dendropemon, Psittacanthus, and Struthanthus, no other extant mistletoe genera compete with Arceuthobium. Competition with epiphytes might be expected in certain tropical situations, but this does not appear to occur because only young shoots are susceptible to infection by dwarf mistletoes and dwarf mistletoes do not occur in moist tropical forests. Thus, newly emerging host shoots could perhaps be infected by dwarf mistletoes before epiphytes become established. Once the endophytic system is developed, the presence of epiphytes is likely immaterial so long as pollination and dispersal are not impeded.
Arceuthobium is likely the most highly specialized of all mistletoe genera. Within the genus, different species clearly exhibit various degrees of host, morphological, and physiological specialization. Because the majority of North American conifer species are parasitized by Arceuthobium, it appears that the basic, adaptive radiation has already taken place, and that current adaptive trends favor more specialized niches. The climax host species were perhaps colonized first by the more primitive species of dwarf mistletoes; then as those ecological niches (host species) were filled, further evolution necessitated the colonization of host species characteristic of pioneering or intermediate stages of the ecological succession. Thus, species (such as A. americanum and A. douglasii) with a relatively large number of derived characteristics—including the formation of systemic infections—parasitize trees that are intermediate in forest succession (Pinus contorta and Pseudotsuga menziesii, respectively).
Two of the most significant factors in the evolution of vascular plants are natural hybridization and polyploidy (Grant 1981). So pervasive are these features that most groups of vascular plants appear to show some evidence of one or both of these phenomena.
In Arceuthobium, however, polyploidy is unknown. In our cumulative years of field experience and that of a number of knowledgeable colleagues, we have encountered no examples of natural hybridization, even though our studies of sympatry (chapter 5) led us to consciously search for such evidence. We have not found any mention of hybridization or polyploidy in the extensive literature on dwarf mistletoes, and experimental crosses between A. apachecum (male parent) and A. blumeri failed to set seed (R. L. Mathiasen, personal communication).
The near absence of hybridization is apparently typical of Loranthaceae and Viscaceae (Barlow and Wiens 1971, Wiens and Barlow 1971). In Phoradendron, a single clear instance of natural hybridization is known (P. densum x P. juniperinum), but the resulting hybrid was apparently sterile (Wiens 1962, Vasek 1966, Wiens and Dedecker 1972). Two examples of hybridization are also reported in Australian Loranthaceae, but these rare situations likewise have little evolutionary importance (Bernhardt 1983). The absence of natural hybridization in Arceuthobium, as well as other mistletoes, can be explained by the occurrence of strong, interspecific, isolating mechanisms, particularly seasonal isolation due to variation in flowering times among sympatric species (Wiens 1964, 1968).
Another factor that might explain the apparent absence of hybrids in Arceuthobium is the lack of suitable habitats for their establishment. Most dwarf mistletoes have a principal host that the species most commonly parasitizes. Because hybrids combine the genetic characteristics of two species, they are often best adapted to intermediate habitats. However, intermediate habitats for hybridized parasites theoretically could be produced only through hybridization of the two host species in question. In other words, "hybridization of the habitat," as Anderson (1948) expressed it, would literally be necessary. Where host pines hybridize, the hybridizing pine species are so closely related that they are usually already parasitized by the same dwarf mistletoe: e.g., A. campylopodum on Pinus jeffreyi x P. ponderosa in California, and A. americanum on Pinus contorta x P. banksiana in Alberta, Canada.
Whether hybrid establishment in dwarf mistletoes is precluded by strong pre-zygotic isolating mechanisms (as we suspect) or because hybrid seeds fail to become established, the result is the same. In either case, there should be strong selective pressures, especially among parasites, against gene combinations that permit hybridization (Dobzhansky 1951).
Absence of natural hybridization (or hybrid establishment) provides a theoretical explanation for the absence of polyploidy in Arceuthobium and other mistletoes (Wiens 1968, 1975). This is based on the assumption that most polyploids are of hybrid origin (alloploids) (Stebbins 1950). If hybridization does not occur, alloploidy is precluded. Autoploidy might still occur, but it is probably of limited evolutionary significance. Several cases of known polyploidy in Viscum might be of autoploid origin, but others are not (Wiens 1975).
In the absence of hybridization and polyploidy, the various species of dwarf mistletoes have apparently maintained distinct phyletic lines that produce the dendritic patterns of evolution typical of animals. In vascular plants, however, reticulate evolution, the product of natural hybridization, is the rule. Alloploidy likely stabilizes natural hybrids, and they evolve into true-breeding species. Hybridization and polyploidy, therefore, result in combining and then stabilizing the characteristics of different evolutionary lines, thereby producing reticulate evolutionary patterns characteristic of vascular plants.
As previously mentioned, a dendritic evolutionary pattern appears to characterize Arceuthobium, but the situation is obscured by the extreme morphological reduction attendant with the parasitic habit. Most species of Arceuthobium are relatively distinct, but the characters separating them are often cryptic or discernible for only short periods of the life cycle. Even though particular characters are not always clearly evident, when the characters are analyzed in their totality, the taxa become well differentiated. This is evident from numerical analyses (Hawksworth and Wiens 1972), which show that all taxa have a relatively high degree of integrity and tend to be more distinct than in other groups where hybridization has played an important evolutionary role (Sokol and Sneath 1963).
The two subgenera or phyletic lines (flabellate or verticillate branching patterns) proposed by Kuijt (1970) are accepted here (see table 14.1 and 15.2) but the extreme reduction of shoots in some species (Arceuthobium pusillum, A. chinense, A. sichuanense, A. tibetense, and A. minutissimum) can obscure the branching patterns. In Arceuthobium, the evolutionary pattern can be generally characterized by three stages of evolutionary divergence that appear to be present in both the New and Old Worlds but are better developed in the former.
Because the evolutionary patterns are dendritic, these evolutionary stages generally correspond to ancestral (pleisiomorphic), intermediate, and highly derived (apomorphic) species groups (chapters 14 and 15). Some characteristics that appear to exhibit evolutionary trends in Arceuthobium are summarized below (adapted and revised from Gill 1935):
| Characteristic | Pleisiomorphic | Apomorphic |
|---|---|---|
| Host specificity | Low | High |
| Shoot size | Large | Small |
| Fruit maturation period | Long | Short |
| Shoot longevity | Long | Short |
| Polymorphism | High | Low |
| Flowering group | Direct | Indirect |
| Branching type | Verticillate | Flabellate |
| Witches’ brooms | Non-systemic | Systemic |
By most of these criteria, Arceuthobium americanum, A. douglasii, A. guatemalense, and A. pusillum in the New World and A. minutissimum in the Old World are highly derived species. Conversely, A. abietis-religiosae, A. globosum, A. verticilliflorum, and A. vaginatum in the New World and A. azoricum and A. pini in the Old World are the least specialized, ancestral species. Intermediate species are well represented in section Campylopoda (chapters 14 and 15). These species groups are not cladistic alliances. They were picked solely to represent species with varying proportions of ancestral and derived characters.
Wiens (1968) studied the flowering characteristics of New World dwarf mistletoes. He hypothesized that the different flowering groups arose as the genus moved northward from Mexico (the present area of greatest species diversity) into latitudes of greater annual variation in photoperiod and temperature. He concluded that species with indirect, spring flowering (Arceuthobium americanum, A. douglasii, and A. pusillum) were derived from species with direct, summer flowering patterns, which in turn were derived from species with direct, spring periods of flowering.
Kuijt (1970) presents convincing evidence that the flabellate (fan-shaped) branching pattern, common in most North American taxa, is a derived character. The more ancestral, verticillate branching is present in all five of the Old World species in which this trait has been studied and in three New World species, Arceuthobium abietis-religiosae, A. americanum, and A. verticilliflorum.
The formation of systemic (isophasic) witches’ brooms (see figs. 6.3 and 16.57) is a derived stage of evolution. Although systemic types of witches’ brooms are occasionally induced by several dwarf mistletoes, they are consistently formed by only five species—Arceuthobium pusillum, A. douglasii, A. guatemalense, and A. americanum in the New World and A. minutissimum in the Himalayas. The type of witches’ brooms formed by the other Old World species in the Himalayas is poorly documented, but apparently A. tibetense and A. chinense may also produce systemic witches’ brooms (Kiu 1984b).
The host–parasite combinations where witches’ brooms are not consistently formed may represent the original, ancestral state. These likely preceded non-systemic, localized formation of witches’ brooms (see fig 6.4), which may in turn have preceded systemic formation of witches’ brooms. Systemic witches’ brooms have adaptive significance because they increase reproductive output of the dwarf mistletoe. By the formation of systemic witches’ brooms, the endophytic system from a single seed can ramify through hundreds of linear meters of host branches and produce a profusion of shoots, flowers, fruits, and seeds.
Most dwarf mistletoes eventually kill their hosts, but mortality occurs much sooner in some host–parasite combinations than in others. This factor, however, is apparently not related to the relative stage of evolutionary development as measured by the characters described above. Within the evolutionarily intermediate group, some species cause severe host mortality (e.g., Arceuthobium laricis on Larix occidentalis and see table 12.5), and other species cause relatively little host mortality (e.g., A. divaricatum on Pinus edulis).