The known world fauna of aphids (Aphidinea) recently reached a total of 5000 species, placed in 510 currently accepted genera. About half of these species spend all or part of their life feeding on trees, and it can be seen from Table 1 that all the major groups of aphids are mostly or even entirely associated with trees. The proportion of tree-living aphid species is probably even higher than indicated, as the unknown hosts of many species are likely to be trees. The trees most favoured as hosts tend to be the older evolutionary groups such as Coniferae, Lauraceae, Fagaceae, Betulaceae, Hamamelidaceae, Ulmaceae and Juglandaceae, and it seems likely that the major groups of aphids differentiated before the appearance of herbaceous plants. Only three groups at or above the tribal level live only on herbs; the Saltusaphidinae which live on Cyperaceae and Juncaceae, the Siphini (subfamily Chaitophorinae) living on Gramineae/Poaceae, and the Tramini (subfamily Lachninae) living mostly on roots of Compositae/Asteraceae.

Nevertheless, aphid species collectively have been recorded from about 300 plant families (Table 2), including many of the more recent families of mainly herbaceous flowering plants. Some of these families (marked by an asterisk in Table 2) only have the most polyphagous aphid species recorded from them, but there are also very numerous host plant-aphid associations at the genus or species level, without which it would have been impossible to have compile the host-oriented keys on this website. The discrepancy is mainly due, as can be seen by comparing aphid species numbers in Table 1, to the massive recent expansion of the largest subfamily Aphidinae. Many members of both tribes within this subfamily (Aphidini and Macrosiphini) retain an association with woody hosts, usually Rosaceae, but migrate for the summer to a wide variety of herbaceous plants, including ferns and mosses as well as angiosperms, and some of the largest genera of Macrosiphini live exclusively on herbaceous plants.

Aphids are predominantly a northern temperate group, with remarkably few species in the tropics. Dixon et al. (1987) postulated that the great diversity of the tropical forest fauna mitigates against short-lived host-specific insects such as aphids. Certainly the absence of aphids from many tropical forest trees is striking, with whole families (e.g. Dipterocarpaceae) seemingly almost immune from attack. The absence of records of aphids from economically important tropical forest trees such as mahogany (Swietenia mahogoni, Meliaceae) and rosewood (Dalbergia nigra, Leguminosae) can hardly be due to negligence by collectors, and suggests that aphids really do not occur on such trees, or at least do little damage. We think, however, that the explanation for this can be found in the evolutionary history of aphids rather than in their present-day host relations or ecology. Psyllids have similar ecology and host relations to aphids, yet many tropical trees with few aphids bear a large psyllid fauna. It seems likely to us that aphids have failed to diversify in the tropics because of one particular, primitive feature of aphid biology, their cyclical parthenogenesis.

Cyclical parthenogenesis is a very successful way of exploiting the short-lived growth flushes of temperate plants, and aphids are thus a very successful group in temperate climates, using seasonal clues to time the alternation of the sexual and parthenogenetic phases of their life cycles (see below). Such life cycles cannot however be readily adapted to tropical conditions. Aphids moving from temperate zones into the tropics simply lose the sexual phase of the life cycle, and in doing so they lose the potential to evolve and diversify that is dependent on the recombination of genes. The tropics may also have acted in this way as a barrier to aphid colonization of southern temperate regions, which also have very small indigenous aphid faunas.

The occurrence of Neophyllaphis on Podocarpus, Araucaria and related conifers throughout the southern continents testifies to the age of aphid-tree relationships, but very little is known about such evolutionarily ancient associations. Most ecological and experimental studies of aphid-tree interactions have concerned introduced species. In Britain, economic damage to spruce by sporadic outbreaks of Elatobium abietinum has been documented since 1846. Damage to the more recently introduced Picea sitchensis is particularly severe, heavy infestations resulting in complete needle loss. Aphid infestations have been shown to reduce the accretion of wood (e.g. of sycamore; Dixon, 1971a), and have deleterious effects on tree root growth (e.g. of Tilia; Dixon, 1971b). However, none of these trees is native to Britain. There are no native British Picea, sycamore is an introduction from Central Europe, and the common British lime tree (or linden) is thought to be a hybrid between a native and an introduced species.

Planted forests of exotic trees cover enormous areas of the globe. There are more than 5.5 million hectares of planted forests in Brazil, of which at least 40 are Eucalyptus spp. (Anon., 1985). Pinus radiata occupies only a small area in its native California but has been widely planted in New Zealand and elsewhere. During this century many European, oriental and American species of Pinus were introduced to various parts of Africa and grew aphid-free for many years. In recent times three aphids, Eulachnus rileyi from Europe, Cinara cronartii from North America and Pineus boerneri of uncertain origin, have appeared on pines in Africa and caused far greater damage than they do in Europe or America. Similarly, Cinara cupressi is much more damaging to Cupressaceae in Africa than in Europe. These exotic conifers may be growing under stress, and the aphids are certainly without the complex of natural enemies associated with them in their countries of origin.

Most aphid damage to trees seems to result directly from feeding, either by removal of sap or wounding of tissue, or in at least some cases by the toxic effect of saliva. Aphids are rarely recorded as vectors of viruses infecting trees (Biddle and Tinsley, 1967). Given the astronomical numbers of aphids in the air and the length of life of trees, there must be strong selection among trees for resistance to aphid-transmitted viruses. It would be interesting to know the mechanism of this resistance, and whether it could be transferred to shorter-lived crop plants. Perhaps the energy required to maintain such defences would be uneconomic for annual or biennial plants.

The problem of how a long-lived plant genotype such as an individual tree survives, when its herbivores have numerous generations in which to evolve methods of breaking its defences, is discussed by Whitham (1983), who showed that there is a similar range of resistance to attack by Pemphigus betae among different branches of one cottonwood tree, as there is among trees in a population. He concluded that long-lived plants are mosaics of phenotypic and/or genetic variability, the genetic differences possibly arising by somatic mutation (Whitham and Slobodchikoff, 1981).

Alstad and Edmunds (1983) found that the black pine-leaf scale, Nuculaspis californica, seemed to establish demes with genetic adaptations to counteract the defence patterns of individual ponderosa pine trees. It is not known whether any aphids develop such long-term natural associations. Tree-dwelling aphids, especially those of the large subfamily Calaphidinae, tend to be rather more active insects than the aphids which colonize herbaceous plants, and may frequently move between trees - although the extent of movement by individual aphids is still largely unknown. Most aphid species in several other subfamilies alternate annually or biennially between their tree host and a herbaceous host (see below), and therefore cannot develop genotype-specific associations, unless of course they were to return to the same tree year after year.



Aphid life cycles can be quite complicated and involve a succession of morphologically different forms (morphs) of the same species. The complexity and the terminology created to describe it - can be daunting to the non-specialist. Rather than add to the pages of descriptive text already available on aphid life cycles (e.g. Hille Ris Lambers 1966d, Blackman 1974, Dixon 1985, Miyazaki 1987), we will merely summarize the essential features, avoiding jargon as much as possible, and use diagrams (Figs 1-7; links below) to illustrate typical life cycles of tree-dwelling aphids. Some unavoidable additional terminology - for example that needed to describe adelgid morphs and life cycles - can be picked up by studying the life cycle diagrams. The essential features of aphid life cycles are:

1. The various families and subfamilies of Aphidoidea each have life cycles with characteristic features, indicating that they have evolved independently.

2. A complete life cycle (that is, a holocycle) typically consists of one generation of sexual morphs (sexuales) and several generations in which only parthenogenetic females are produced. This phenomenon of cyclical parthenogenesis is a basic, primitive feature of aphid biology.

3. In the more primitive families, Adelgidae and Phylloxeridae, both sexual and parthenogenetic females are oviparous, but in the Aphididae parthenogenetic females always give birth to live young; in Aphididae the parthenogenetic females are therefore termed viviparae, and the sexual females are distinguished as oviparae.

4. The more complex life cycles involve host alternation; the technical term is heteroecy. In heteroecious aphids, the sexuales mate and fertilized eggs are laid on a tree or shrub, the primary host, but a regular migration occurs at some stage in the life cycle to another, totally unrelated plant, which may be herbaceous or woody - the secondary host. On the secondary host, only parthenogenetic generations (exules) occur, and a return migration to the primary host is needed before the next sexual generation.

5. Because host alternation has evolved several times independently in Aphidoidea, there are important differences at the family and subfamily levels in the way in which it is achieved (see Figs 1-4, links below). It may occur as part of a one-year cycle (this happens in all Aphidinae, most Hormaphidinae, Pemphigini, Eriosomatini), or the complete cycle may take two years (Adelgidae, Fig. 1; Fordini, Fig. 2). Some Hormaphidinae have now been shown to have long-lasting galls on their primary hosts, that do not mature for 2-5 years.

6. The great majority of aphids go through both the sexual and parthenogenetic phases of their life cycle on one host plant, or on a small range of closely-related plants. The technical term for this is monoecy. Calaphidinae, Drepanosiphinae, Chaitophorinae, Greenideinae and Lachninae do not have host alternation; all species in these subfamilies are monoecious. Some examples of monoecious life cycles are depicted in Figs 5, 6 and 7 (links below). Monoecious aphids generally have fewer morphs, and there are smaller differences between morphs, than in heteroecious aphids, although there may be considerable seasonal variation.

7. Some aphids have lost the sexual part of the life cycle; that is, they are anholocyclic. Some species are entirely anholocyclic and have no known sexual morphs (e.g. Tuberolachnus salignus, Pineus boerneri, Myzus ascalonicus), while others may be anholocyclic in warmer climates and holocyclic in cold temperate regions (e.g. Eulachnus rileyi, Myzus persicae). Populations of certain species maintain the options of both sexual and parthenogenetic reproduction in mild climates, by producing sexuales while at the same time continuing to produce parthenogenetic females (e.g. many Greenideinae). Anholocyclic populations of heteroecious aphids lose their link with the primary host and live all year reproducing parthenogenetically on secondary host plants.


Figs 1-7 Life cycle diagrams

Fig. 1 Adelges laricis representative of two-year cycle of Adelgidae

Fig. 2 Baizongia pistaciae representative of two-year cycle of Fordini

Fig. 3 Thecabius affinis typical cycle of heteroecious Eriosomatinae

Fig. 4 Aphis fabae typical cycle of heteroecious Aphidinae

Fig. 5 Periphyllus testudinaceus with aestivating first instar morph on Acer

Fig. 6 Drepanosiphum platanoidis with aestivating alatae on Acer

Fig. 7 Cinara schwarzii typical cycle of Cinarini