Elatobium abietinum (green spruce aphid)

Elatobium is a small genus of five biologically diverse, tree-feeding species within the very large aphid family Aphididae. Besides E. abietinum, which feeds on spruce (Picea spp.) in Europe and other parts of the world, the genus includes Elatobium hidaense on Salix in Japan and Kamschatica, Elatobium laricis on Larix sibirica in east Siberia, Elatobium momii on Abies firma, Picea jezoensis and Taxus cuspidata in Japan, and Elatobium trochodendri on Trochodendron aralioides in Japan (Blackman and Eastop, 1994; Carter and Halldórsson, 1998). The Japanese species were revised by Miyazaki (1971), although his concept of the genus was broader and included related aphids on Ericaceae that are now placed in the genera Ericolophium and Neoacyrthosiphon (Blackman and Eastop, 1994).E. abietinum was first described as Aphis abietina by Francis Walker in 1849, from material collected on Norway spruce (Picea abies) in London, UK. At this time, many aphid species were being assigned to the genus Aphis, but detailed drawings of the lectotype in the collection of the Hope Department of Entomology, Oxford, UK, by Doncaster (1961) provide little doubt that this is the same species as E. abietinum (Carter and Halldórsson, 1998). Theobald (1926) proposed the new generic name Neomyzaphis, and the combination Neomyzaphis abietina remained in popular usage for many years. However, Mordvilko (1914) had assigned the generic name Elatobium to the species at an earlier date, which therefore takes precedence over Neomyzaphis. However, the name Elatobium remained largely unused until the publication of works by Cottier (1953), Doncaster (1961) and Eastop (1966).

E. abietinum is a serious invasive pest in North America where susceptible spruce species form a major component of the native forests. Defoliation of Picea sitchensis along the Pacific coast, and of Picea englemannii, Picea glauca and Picea pungens in the south-west, has the potential to alter forest succession and species composition, and to impact on other flora and fauna. The highly dispersive nature of the alate aphids during the flight period means that once established, E. abietinum will eventually spread throughout the forested area to the limits of its climatic tolerances.In other parts of the world where E. abietinum is introduced, both the aphid and spruce are exotic and there is no risk of invasion into natural habitats. However, the mobility of the winged aphids and potential for rapid population increase once a stand has been colonised, means that all trees must be considered likely to be attacked at some point during their life.

Detailed descriptions and illustrations of the nymphs and adults of apterous and alate virginoparae are provided by Cunliffe (1924), Dumbleton (1932), Cottier (1953), von Scheller (1963) and Heie (1992). The sexual stages (ovipara, alate male, eggs and fundatrix) are described and illustrated by von Scheller (1963).
Apterous Viviparous Female

Body green, head yellowish-green with dull-red eyes. Rostrum pale-green, black on apex. Antennae yellowish-green, distal segments (III-V) can be dusky, apical segment (VI) black. Cauda pale-green to yellowish-green. Legs pale-greenish, femora and tibia dusky distally, tarsi black. Body length 1.0-1.6 mm (Cunliffe, 1924; Dumbleton, 1932; Cottier, 1953), 1.4-2.0 mm (Heie, 1992). Antenna six-segmented with large circular or oval sensilla, 0.45-0.55 x body. Rostrum reaching to (Cottier, 1953) or behind hind coxae (Heie, 1992). Mean length of antennal segments: I, 0.07; II, 0.05; III, 0.21; IV, 0.14; V, 0.14; base VI, 0.11; flg. VI, 0.16 mm. Mean ratio: length of antennae / length of insect = 0.6. Body without marginal tubercles. Cornicles cylindrical, tapering only slightly from base to apex, which has a conspicuous lip. Mean length, 0.39 mm. Cauda spinose, elongate, constricted at middle. Mean length, 0.18 mm.

Apterous Nymphs

Green, oval. Four nymphal instars. Antennae four-segmented in the first-instar, five-segmented in the second- and third-instars, and six-segmented in the fourth-instar. Mean sizes (length x width): first-instar, 0.6 x 0.2 mm; second-instar, 0.9 x 0.3 mm; third-instar, 1.0 x 0.3 mm; fourth-instar, 1.2 x 0.4 mm (Cunliffe, 1924).

Alate Viviparous Female

Head green. Three ocelli present. Antenna six-segmented with large circular or oval sensilla, 0.6-0.7 x body. Mean lengths of segments: I, 0.07; II, 0.05; III, 0.35; IV, 0.23; V, 0.20; base VI, 0.12; flg. VI, 0.17 mm. Mean ratio: length of antennae / length of insect = 0.7 (Cottier, 1953). Pronotum light-green with darker markings. Pre-scutum, scutum, scutellum and post-scutellum dark-brown to black. Metanotum dark-green. Mesosternite dark-green to black. Legs with tarsi and tips of tibiae darker. Abdomen green, with intersegmental muscle sclerites; antennae and siphunculi brownish. Body without marginal tubercles. Cornicles dusky, rather thin, almost same width from base to apex. Mean length, 0.43 mm. Cauda pale-green, spinose, usually constricted at basal third to middle, and somewhat pointed. Mean length, 0.18 mm. Rostrum reaching to middle coxae or longer. Body length 1.0-1.8 mm (Dumbleton, 1932; Cottier, 1953), or 1.6-2.1 mm (Heie, 1992). Wings much longer than body, venation normal.

Carter (1969) describes an intermediate alate/apterous form of E. abietinum collected during the winter in Scotland, which probably developed because of exceptionally mild weather.

Alate Nymphs

Instars I-III similar to apterae. Fourth-instar with prominent dorsal wing buds (von Scheller, 1963).

Oviparous Female

Rostrum reaching to abdominal segment III. Body length approximately 1.3 mm. Otherwise much like the apterous viviparous female (von Scheller, 1963; Heie, 1992).

Alate Male

Similar to alate virginoparae, with three ocelli. Body length approximately 1.5 mm. Abdomen with rather pale and narrow dorsal cross bars on tergites III-VIII and marginal spots. Antennae as long as body, with larger numbers of sensilla (von Scheller, 1963; Heie, 1992).

Eggs

Yellow to reddish at first, turning dark-brown or black. Oval, 0.65 x 0.30 mm. Illustrated by von Scheller (1963). Unfertilised eggs remain yellow and eventually shrivel and die.

Fundatrix

Similar to apterous virginoparae (von Scheller, 1963).

E. abietinum is considered native to Europe on Picea abies and introduced elsewhere (Kloft et al., 1961, 1964; Carter and Halldórsson, 1998). In maritime countries of north-west Europe, where there are no native spruce species (e.g. Ireland, UK, Denmark, north-west France) E. abietinum has spread naturally to colonise P. abies, Picea sitchensis and other Picea spp. in all types of situations. In other parts of the world, the aphid has been introduced. In addition to the distribution list of this datasheet, E. abietinum is widespread and exotic in Belgium (J-C Gregoire, University of Brussels, Belgium, personal communication, 2004). In New Zealand, Tasmania, Iceland, Chile and other regions where there are no native spruces, the aphid is confined to commercial plantations of spruce, Christmas tree crops, spruce shelterbelts and ornamental plantings. In North America, in contrast, introduction has led to the establishment of E. abietinum in natural forests of P. sitchensis along the Pacific coast and in Picea englemannii / Picea pungens forests in the interior south-west. The aphid has spread or is spreading through large areas of these natural forests.

For many geographically isolated parts of Europe, notably the UK and Iceland, there are no native spruce trees. Although Picea abies is thought to have been cultivated in Britain from 1548 and Picea sitchensis from 1832, it was not until the two post-war decades 1945 to 1960 that really extensive afforestation with these species took place (Carter and Halldórsson, 1998). Records of E. abietinum in Britain by the time of Theobald (1926) suggested that the aphid was already well-distributed over much of the country. The raising of planting stock in areas where infestation of the young trees could readily take place and then transporting these to remote planting sites, no doubt accelerated the spread of the aphid in the UK, and the same circumstances probably apply to other countries (Carter and Halldórsson, 1998).In France, there was an increase in the planting of P. sitchensis between 1950 and 1960, especially in coastal areas of Brittany (Leroy and Malphettes, 1969). The widespread outbreak of E. abietinum in 1961 that affected much of Europe extended into these young plantations, and appeared to be especially damaging on poor shallow soils (Joly, 1961; Leroy and Malpettes, 1969). Since this time, the area of P. sitchensis plantations in France has increased and is now in excess of 30,000 ha. P. sitchensis plantations occur as far south as 45°N on the Millevache plateau in Limousin. E. abietinum is well distributed in these areas, but outbreaks in 1987 and 1988 were only recorded from Brittany. These outbreaks occurred on poor sites and mostly in 30- to 35-year-old stands. In Limousin, the last outbreak was recorded in 1994 near Limoges between 500 and 600 m altitude (Carter and Halldórsson, 1998).E. abietinum is common in Belgium (J-C Gregoire, University of Brussels, Belgium, personal communication, 2004), and would be expected to occur in Luxemburg, but there are no literature records for its presence in either of these countries. E. abietinum was first recorded in Iceland in 1959, in a nursery in Reykjavik. At the time, the nursery was used for storing imported Christmas trees. Attempts to eradicate the aphids failed and it was found in several gardens in Reykjavik in late summer and autumn 1960 (Bjarnason, 1961; Ragnarsson, 1962). During the 1960s and 1970s, the aphid spread eastwards and colonised stands of P. sitchensis and other North American spruces in south Iceland. During the 1980s, it colonised stands in east Iceland, and during the 1990s it spread west and north. E. abietinum has now been found in almost all spruce stands in the country (Ottósson, 1985; Carter and Halldórsson, 1998; Halldórsson et al., 2003). The first major outbreak of E. abietinum in Iceland occurred in the south-west in 1964. Since then, another five major outbreaks have occurred, in 1977, 1984, 1987, 1991 and 1996 (Ottósson, 1985; Blöndal, 1988). All outbreaks in Iceland have been in the late summer / autumn / early winter period, and outbreaks have never occurred after severe winters (Halldórsson, 1995).Analyses of genetic variation suggest that E. abietinum has probably been introduced into Iceland on more than one occasion (Sigurdsson et al., 1999). Icelandic aphids are genetically most closely related to aphids from Denmark, which agrees with their assumed origin (Sigurdsson et al., 1999; Halldórsson et al., 2004). Christmas trees were imported annually from Denmark into Reykjavik from 1959 up to 1990 (Sigurdsson et al., 1999). The first report of E. abietinum in North America was on P. sitchensis in Vancouver, British Columbia, in 1914 (Kloft et al., 1964). Extensive damage to P. sitchensis has been reported in British Columbia since 1916, and E. abietinum is now known throughout the coastal range of P. sitchensis from California to Alaska (Keen, 1952; Pillsbury, 1960; Holms and Ruth, 1968, Koot, 1991). Infestations are restricted to the coast and none have been reported inland (Koot, 1991). Severe infestations have occurred in coastal areas of the Queen Charlotte Islands since 1960, and between 1981 and 1991 there was significant damage to both natural spruce forests and ornamental trees (Koot, 1991). In most areas, E. abietinum has been a chronic pest, and outbreaks have been sporadic and difficult to predict (Wood et al., 1985; Lewis et al., 1999). However, severe defoliation has caused mortality in local areas, and in some stands up to 67% of the trees have been killed (Wood et al., 1985; Koot, 1991; Van Sickle, 1995). Seed orchards of P. sitchensis and Picea glauca have also been affected (Wood et al., 1985; Lewis et al., 1999). A small number of E. abietinum were collected from P. glauca in Quebec in November 1962 (Kloft et al., 1964). However, the aphid could not be found in the following year and there were no further records of E. abietinum east of the Rocky Mountains until Fedde (1971, 1972) discovered the species on P. glauca in North Carolina. There were no further reports of E. abietinum in eastern North America until 2002, when severe infestations on P. glauca were reported from coastal areas of Rhode Island (mainland Rhode Island and Block Island) and Massachusetts (Cape Cod, Martha's Vineyard) (Anon., 2002).In the interior south-west USA, E. abietinum was first found in urban Sante Fe, New Mexico, in 1976, where it has remained as an intermittent pest in the urban forest (Lynch, 2002, 2004). The first wildland outbreak in the south-west occurred over the 1989-1990 winter in the White Mountains of Arizona, when more than 100,000 acres of Picea englemannii and Picea pungens were defoliated (Lynch, 2004). These two spruce species occur naturally in high elevation, mixed-conifer forests above 2400 m in the mountains of Arizona, New Mexico and Utah, and are highly susceptible to E. abietinum. Three further outbreaks of E. abietinum occurred over the winters of 1995-1996, 1996-1997 and 1999-2000 (Lynch, 2004). The range of E. abietinum has now expanded to include the Mogollon Mountains (east of the White Mountains) and Sacramento Mountains in New Mexico, and the Pinaleno Mountains and San Francisco Peaks in Arizona. Outbreaks appear to be associated with dry winter and spring weather prior to the autumn and winter in which feeding occurs (Lynch, 2002). Despite the high elevations and potentially very cold climate, evidence to date suggests that all of the E. abietinum populations are behaving anholocyclically. The origin of the aphids present in Arizona, New Mexico and Utah is not known.The earliest record of E. abietinum in New Zealand is that of Myers (1922), who stated that it had been killing spruce in large quantities at Taranaki (Dumbleton, 1932). Infestations in 1920 appear to have reached epidemic proportions. The aphid is now found throughout the North and South Islands, on P. abies, P. sitchensis and Picea smithiana (Dumbleton, 1932; Rawlings, 1953; Cottier, 1953; Zondag, 1983). Defoliation by E. abietinum appears to have been a contributory cause to the unhealthy condition of spruce in New Zealand up to 1929 (Dumbleton, 1932), and the aphid is still considered to be a major factor preventing spruce being used as a production forest species in New Zealand (Miller and Knowles, 1989; Nicol et al., 1998).In Australia, E. abietinum is only known from Tasmania, where it was first recorded in 1943 (Evans, 1943).

Due to the impact that it has had since becoming established in New Zealand and North America, E. abietinum should be of concern to those countries with plantations or natural forests containing susceptible spruce species where the aphid has not yet been introduced.

Host tree species of E. abietinum are restricted to the genus Picea (Hanson, 1951). Feeding has been reported to occur on species in other conifer genera, e.g. Abies spp., Larix sibirica, Pinus strobus, Pinus sylvestris, Pseudotsuga menziesii (Theobald, 1914; Hussey, 1952; Volkova, 1970; Szelegiewicz, 1975), but this is exceptional and may only occur during outbreaks. The number of aphids involved is usually small and the damage caused unimportant (Furniss and Carolin, 1977; Blackman and Eastop, 1994; Carter and Halldórsson, 1998). Attempts to rear E. abietinum on hosts other than Picea under laboratory conditions have usually failed (e.g. Cunliffe, 1924). The many distribution records of E. abietinum throughout Europe, and the existence of holocyclic populations in central Europe, suggest that the native host is Picea abies (Theobald, 1914; Kloft et al., 1961, 1964; Bejer-Petersen, 1962; Carter and Halldórsson, 1998). This is supported by the fact that P. abies and other Palaearctic spruce species show less reaction to the aphid's feeding than does Picea sitchensis and other Nearctic species. The other aphid species assigned to the genus Elatobium are all from Asia, and it is also in east Asia that speciation of the genus Picea appears to have taken place. Possible exposure to an Elatobium species in this region may be the reason why the most highly resistant Picea species occur there (Carter and Halldórsson, 1998).Theobald (1914, 1926) listed 14 species of Picea as food plants and recognised from observations at Kew (London, UK) that European and Asiatic species were generally damaged much less than North American species. More detailed studies on the relative susceptibility of 20 different Picea species to E. abietinum were carried out by Nichols (1987). Field observations showed that Picea species could be assigned to two distinct groups. The first group consisted of the North American species (except Picea rubens) which all supported high populations. Picea sitchensis, Picea pungens, Picea englemannii and Picea glauca were particularly susceptible. The second group contained all the Eurasian, Chinese and Japanese spruces (except Picea asperata) and supported the least number of aphids. Laboratory studies on the performance of individual aphids indicated that Eurasian spruces are intermediate in terms of resistance between North American and Asian species. The mean relative growth rate (MRGR) of individual E. abietinum on P. abies is similar to that on P. sitchensis (Fisher, 1987; Nichols, 1987), but P. abies loses fewer needles in response to attack (Parry, 1974a; Fisher, 1987; Straw and Green, 2001).The exceptions to this general pattern of resistance and susceptibility are P. rubens and Picea breweriana, North American species that are highly resistant to attack, and Picea asperata from China that is highly susceptible (Nichols, 1987). The mechanisms underlying resistance to E. abietinum are not known, but may include variation in secondary compounds, particularly terpenes, availability of sap nutrients and anatomical differences in needle structure (Harding et al., 1998).

Attack by E. abietinum results in the formation of a small yellow spot around the feeding site or band across the needle. These symptoms are characteristic and lead to the whole needle eventually turning yellow and falling from the shoot within a few weeks. Yellow spots around the feeding site develop after about 5-10 days on Picea sitchensis, but longer on Picea abies (Kloft and Ehrhardt, 1959; Parry, 1971, 1974a; Fisher, 1987). The chlorosis appears to be caused by toxic substances, possibly certain amino acids, which are injected into the needle tissues with the aphid's saliva (Kloft and Ehrhardt, 1959). Adjacent needles that have not been fed on, remain healthy and are not necessarily shed.High densities of E. abietinum can lead to severe defoliation, but in spring and summer, because the aphid avoids the current foliage, needle loss is restricted to one-year-old and older shoots. In the autumn and winter, the aphid feeds on current as well as older needles, and high populations at this time are capable of causing complete defoliation. Attack on small P. sitchensis trees (up to 5 years old) is usually concentrated in the upper part of the canopy (Straw et al., 1998a, b), whereas on older trees, attack is usually concentrated in the lower, more shaded portions of the crown (Bevan, 1966; Parry, 1969a; Koot, 1991). Young trees up to 4 years old that have been heavily defoliated during the winter often have terminal buds that fail to break the following spring (Carter, 1977).

Conifer spider mite, Oligonychus ununguis also causes discoloration, death and needle fall on spruce in mainland Europe, UK, and New Zealand, and damage caused by the mite has been confused with that caused by E. abietinum in the past (Dumbleton, 1932; von Scheller, 1963). However, mite damage is more frequent on small trees in dry situations, and on Christmas trees (Carter and Winter, 1998). The mites suck sap from the epidermal and cortical cells, and damaged needles have a yellow speckled (not banded) appearance at first. Later the foliage turns bronze. This change develops from the centre of the tree's crown outward, and affects needles on the current shoots last. It is most noticeable in late summer or autumn (UK). Severely damaged needles may fall from the shoot. Close examination with a hand-lens or dissecting microscope will reveal tiny orange-brown to grey spider mites (0.2-0.5mm) and orange-brown to reddish eggs on the shoots. The mites spin silk webbing across the needles, but this is only obviously visible when very large numbers are present. The mite overwinters as eggs on the current shoots, mainly clustered on the bark close to the terminal bud. The eggs hatch in April to mid-May (in the UK) and damage develops as populations increase through the summer. O. ununguis also occurs on many other conifer species (Strouts and Winter, 1994; Carter and Winter, 1998).Eriophyiid needle mites (Nalepella haarlovi) can produce similar damage to O. ununguis on Picea abies and Picea sitchensis (UK), but the mites are much smaller and do not produce silk. P. abies-affected foliage appears greyish at first but then turns reddish-brown. The discoloration begins in the centre of the tree and then spreads outward. Severe infestation can lead to defoliation of the older needles. Close examination with a x10 magnification lens may reveal pale-orange spherical eggs, about twice the diameter of the silver stomata, on the needles. The elongate mites are amber and two to three times the size of the eggs, but all stages are very difficult to see in the field, and examination with a low power microscope is necessary to confirm diagnosis (Carter and Winter, 1998).Chrysomyxa needle rust can produce chlorotic symptoms similar to those caused by E. abietinum (in mainland Europe and the UK), but the rust causes disconcertingly bright and conspicuous needle discoloration. Infection results in needles with bright-orange or yellow transverse bands, or completely discoloured needles that contrast strongly with adjacent healthy green needles. Needles infected with Chrysomyxa abietis (spruce needle rust) over the autumn and winter produce orange, elongated blister-like pustules on the underside in the spring, which release spores that infect newly developing needles; the spore-bearing needles then fall. Chrysomyxa rhododendri overwinters on rhododendron. Spores produced in the spring infect developing spruce needles. The needles produce characteristic white outgrowths on their underside in late summer. These release orange spores that re-infect rhododendron. The infected needles then fall (Strouts and Winter, 1994)Late spring frost damage can cause needle browning and shoot death, but damage is confined to the new, recently flushed foliage. Lime- or chalk-induced chlorosis can lead to foliage yellowing, needle loss and poor growth in spruce, but in such cases discoloration is distributed generally throughout the canopy, and needles turn completely pale or yellow (and are often unusually small) and lack the distinctive yellow spots and bands typical of E. abietinum attack. Signs of infestation (individual insects, exuviae, cadavers, honeydew and sooty moulds) are absent (Strouts and Winter, 1994).

E. abietinum is able to disperse widely in those regions where it is present, through the activity of the alate aphids during the flight period; all spruce trees in the region have the potential to be colonised. The aphid does not show strict habitat requirements, and can attack trees of all ages, and occurs throughout its range in commercial plantations, shelterbelts and on individual trees planted in parks and gardens. Picea pungens,especially var. glauca, planted as an ornamental tree, is particularly prone to attack. E. abietinum has pest status in forest nurseries and on Picea abies grown as Christmas trees (Carter and Winter, 1998). In North America, the aphid has become abundant in natural forests of Picea sitchensis along the Pacific coast, and in natural mixed-conifer forests at high elevations in Arizona and New Mexico (Lynch, 2004).

E. abietinum is attacked by a wide range of generalist aphid predators, that includes ladybirds (Coccinellidae), hoverflies (Syrphidae), lacewings (Neuroptera), birds (tits, Parus spp. and goldcrest, Regulus regulus) and even bats (Theobald, 1914; Hussey, 1952; Börner and Heinze, 1957; von Scheller, 1958, 1963; Ohnesorge, 1959; Bejer-Petersen, 1962; Austarå et al., 1997, 1998). However, the number of actual species of predator that can be confirmed as feeding on E. abietinum is relatively small, because the majority of records from the field refer only to particular species or types of predator being abundant on heavily aphid-infested trees, and not necessarily feeding on the aphids, and few predators have been the subject of feeding trials under laboratory conditions. There is also very little information on the impact of natural enemies on E. abietinum populations in the field. The only quantitative assessment of impact is by Leather and Owuor (1996) and this is rather limited in scope. However, modelling studies by Crute and Day (1990) and Leather and Kidd (1998) suggest that larval coccinellids, syrphid flies and perhaps also hemerobiids are possibly important in suppressing E. abietinum populations in late summer and autumn.The most important coccinellid predator of E. abietinum in spruce plantations in the UK is the larch ladybird, Aphidecta obliterata. This species is a conifer specialist, and the adults and larvae also feed on Adelges spp. on Douglas-fir, and aphids on other conifers (von Scheller, 1963; Parry, 1980). Brown and Clark (1959), Amman (1966), Parry (1980, 1992) and Leather and Kidd (1998) describe aspects of its biology and ecology. Other coccinellids that have frequently been found on P. sitchensis infested with E. abietinum in Europe are adults and larvae of Anatis ocellata, Coccinella septempunctata, Coccinella undecimpunctata, Myzia oblongoguttata, Adalia decempunctata and Adalia bipunctata (von Scheller, 1958, 1963; Bejer-Petersen, 1962; Austarå et al., 1997). Leather and Owuor (1996) give consumption rates of A. bipunctata feeding on E. abietinum on P. abies.Other beetle predators of E. abietinum include species of Cantharidae, several of which are known to feed on aphids (Frazer, 1988). Von Scheller (1958) found three Cantharis spp. commonly with E. abietinum, and Rhagonycha lignosa has been recorded in great numbers from E. abietinum-attacked trees in the UK (Austarå et al., 1998). Von Scheller (1958) also records the click beetle Athous subfuscus as feeding on E. abietinum.Lacewing larvae, particularly larvae of brown lacewings (Hemerobiidae) and larvae of syrphid flies were the most abundant aphid predators in the canopy of P. sitchensis in Northern Ireland, UK, especially when sampling was conducted at night when the larvae were most active (Crute and Day, 1990). The main lacewing species was Hemerobius micans. Larvae of Hemerobius atrifrons have been observed feeding voraciously on E. abietinum in Norway (Austarå et al., 1998). Spiders (Araneidae) have been considered as very likely contributors to E. abietinum mortality due to high numbers of aphid cadavers observed in webs (von Scheller, 1958; Bejer-Petersen, 1962), but neither species nor quantifications of impact have been given. Harvestmen (Opiliones) have also been suggested as predators (Theobald, 1914).Records of parasitoids of E. abietinum are much less frequent than records of predators, and are restricted to Europe. Hanson (1951) mentions the importance of parasitoids in controlling E. abietinum, and Hussey (1952) observed up to 15% parasitism of E. abietinum populations during late summer. However, other authors report much lower parasitism rates and, generally, it appears that parasitism rarely exceeds 10% (von Scheller, 1963; Parry, 1969a; Leather and Kidd, 1998). Several species of hymenopteran parasitoids have been reared from mummies of E. abietinum: Ephedrus koponeni (Halme, 1992), Lysaphidus schimitsheki (von Scheller, 1963; Mackauer and Stary, 1967; Stary, 1973), Praon flavinode (Fulmek, in von Scheller, 1963), and several hyperparasitoid species (Austarå et al., 1997, 1998). E. koponeni appears to be specialised on E. abietinum, but has only been recorded on E. abietinum in Finland (Halme, 1992).E. abietinum is killed by a number of insect pathogenic fungi, all from Zygomycotina Entomophthorales. Day (1984b, 1986) recorded significant mortality in Northern Ireland caused by Entomophthora planchoniana, and Austarå et al. (1997, 1998) recorded this species and five other species from various locations in the UK, Denmark, Norway and Iceland. Neozygites fresenii was found in all countries, and in Norway and Iceland it was the only pathogen isolated. In Iceland, infection varied between 0 and 19.8% amongst aphids collected in October and November 1994, whereas in the following month only two out of 3577 aphids were infected (Austarå et al., 1998). In contrast, in Denmark and the UK, infection levels never exceeded a few percent. E. planchoniana, N. fresenii and two of the other pathogens isolated from E. abietinum (Conidiobolus obscurus and Erynia neoaphidis) are common pathogens of aphids in annual crops, and the fifth species (Conidiobolus coronatus [Boudierella coronata]) is a generalist pathogen found attacking a range of insect species (Austarå et al., 1998). Zoophthora phalloides, which was recorded from E. abietinum in Denmark, is also found in annual cropping systems, but is relatively rare (Keller, 1991). This species may be more specifically adapted to forest ecosystems (Austarå et al., 1998).

The importance of E. abietinum as a forest pest in Europe and most other areas lies in its ability to cause extensive defoliation, especially of Picea sitchensis, and the effect that defoliation may have on shoot growth and dry matter production. Severe defoliation of P. sitchensis during the spring or early summer reduces height increment in the year of attack by 20-60%, and further reductions in height increment may occur in the following and subsequent years (Carter, 1977; Carter and Nichols, 1988; Seaby and Mowat, 1993; Thomas and Miller, 1994). The reduction in diameter increment may be similar after heavy defoliation (Thomas and Miller, 1994). Orlund and Austarå (1996) found that severe defoliation of 34-year-old P. sitchensis in Norway reduced diameter increments for 7-8 years. The mean reduction in diameter increment over this period was 18.5% for trees that were 15 cm diameter at breast height (dbh), and 40.5% for trees that were 20 cm dbh. The maximum reduction in diameter increment occurred 4-5 years after attack, and increments had still not returned to normal by the end of the study period. The impact of moderate defoliation by E. abietinum in the spring on 4-year-old P. sitchensis has been described by Straw et al. (1998b, 2000, 2002). Height increment of defoliated trees was reduced by 22% in the year of attack, by 11% in the following year, and by 6% in the third year, when compared with undefoliated trees. At the end of the third year, defoliated trees were, on average, 10% shorter than undefoliated trees. Stem diameter and volume increments were not reduced in the year in which defoliation took place, but were reduced in the following year, by 12% and 24%, respectively (Straw et al., 2000). There was no reduction in diameter and volume increments in the third year. The immediate reduction in height increment in the year of attack was caused by a direct effect of aphids on shoot extension growth in the spring, whereas the decrease in diameter and volume increment in the following year was caused by a delayed reduction in needle size (Straw et al., 2002).Defoliation during the autumn may have a greater impact than defoliation during the spring (Carter, 1977). However, the only quantitative estimates of impact following autumn defoliation are those given by Halldórsson et al. (2003) from studies in Iceland. Moderate defoliation of P. sitchensis at one site in south-east Iceland during 1986-1987 and 1990-1991 reduced total height and diameter growth over the period 1987-2000 by 16% and 17%, respectively. Severe defoliation reduced both height and diameter growth by 37%. By 2000, when the trees were 40 years old, trees that had been lightly, moderately and severely defoliated in 1991 were, on average, 9.0, 8.1 and 7.2 metres tall, respectively.In Europe, E. abietinum rarely kills trees and its main effect is to reduce increment. The inability of E. abietinum to attack the new needles before the autumn (see Development and Physiology section under Biology and Ecology), means that trees always carry some foliage after spring or early summer infestation, or soon regain foliage after autumn or winter attack. This is usually enough to ensure that the trees survive and recover. The death of large or small trees may occur under exceptional conditions, e.g. after complete defoliation in autumn and winter when the trees have been left bare for several months, or when spring attack coincides with a severe frost that kills the new growth, but in Europe this is a rare phenomenon (Bevan, 1966; Straw et al., 1998a). The risk of mortality is higher for small, not fully established trees on difficult sites and when defoliation is followed by drought stress.In North America and New Zealand, defoliation by E. abietinum has caused significant tree mortality. The large number of Picea englemannii and Picea pungens trees killed by E. abietinum in the interior south-west USA is likely to have a significant economic impact, although as yet this impact has not been quantified. Damage by E. abietinum has been considered to be a major factor preventing spruce being used as a production species in New Zealand, limiting options for commercial forestry and potential yields (Miller and Knowles, 1989; Nicol et al., 1998).E. abietinum can defoliate spruce over considerable areas. Carter (1972, 1977) indicated that 37,700 ha of P. sitchensis were defoliated by E. abietinum in western parts of Britain in 1971, which was equivalent to 15% of the total area planted with P. sitchensis at the time. Koot (1991) reported that more than 5000 ha of P. sitchensis were defoliated by E. abietinum on the west coast of Canada in 1981, and Lynch (2002) estimated that E. abietinum defoliated 57,000 ha of P. englemannii and P. pungens in Arizona in 1999-2000. Large areas of P. sitchensis and Picea abies have been defoliated in Europe in other years (e.g. in 1957), but the actual area defoliated or number of trees affected were not estimated at the time (Carter and Halldórsson, 1998). Variation in defoliation by E. abietinum has been related to differences in local climate and type of planting (Ohnesorge, 1961), elevation (Joly, 1961) and soils (Leroy and Malphettes, 1969).E. abietinum has pest status in the production of P. abies Christmas trees in Europe, and in the growing of spruce trees for ornamental purposes, particularly blue spruce (P. pungens var. glauca). Seed orchards of P. glauca in British Columbia are also frequently defoliated (Harding et al., 1998). Aphid control in these situations will also carry a significant economic cost.

The importance of E. abietinum as a forest pest in Europe and most other areas lies in its ability to cause extensive defoliation, especially of Picea sitchensis, and the effect that defoliation may have on shoot growth and dry matter production. Severe defoliation of P. sitchensis during the spring or early summer reduces height increment in the year of attack by 20-60%, and further reductions in height increment may occur in the following and subsequent years (Carter, 1977; Carter and Nichols, 1988; Seaby and Mowat, 1993; Thomas and Miller, 1994). The reduction in diameter increment may be similar after heavy defoliation (Thomas and Miller, 1994). Orlund and Austarå (1996) found that severe defoliation of 34-year-old P. sitchensis in Norway reduced diameter increments for 7-8 years. The mean reduction in diameter increment over this period was 18.5% for trees that were 15 cm diameter at breast height (dbh), and 40.5% for trees that were 20 cm dbh. The maximum reduction in diameter increment occurred 4-5 years after attack, and increments had still not returned to normal by the end of the study period. The impact of moderate defoliation by E. abietinum in the spring on 4-year-old P. sitchensis has been described by Straw et al. (1998b, 2000, 2002). Height increment of defoliated trees was reduced by 22% in the year of attack, by 11% in the following year, and by 6% in the third year, when compared with undefoliated trees. At the end of the third year, defoliated trees were, on average, 10% shorter than undefoliated trees. Stem diameter and volume increments were not reduced in the year in which defoliation took place, but were reduced in the following year, by 12% and 24%, respectively (Straw et al., 2000). There was no reduction in diameter and volume increments in the third year. The immediate reduction in height increment in the year of attack was caused by a direct effect of aphids on shoot extension growth in the spring, whereas the decrease in diameter and volume increment in the following year was caused by a delayed reduction in needle size (Straw et al., 2002).Defoliation during the autumn may have a greater impact than defoliation during the spring (Carter, 1977). However, the only quantitative estimates of impact following autumn defoliation are those given by Halldórsson et al. (2003) from studies in Iceland. Moderate defoliation of P. sitchensis at one site in south-east Iceland during 1986-1987 and 1990-1991 reduced total height and diameter growth over the period 1987-2000 by 16% and 17%, respectively. Severe defoliation reduced both height and diameter growth by 37%. By 2000, when the trees were 40 years old, trees that had been lightly, moderately and severely defoliated in 1991 were, on average, 9.0, 8.1 and 7.2 metres tall, respectively.In Europe, E. abietinum rarely kills trees and its main effect is to reduce increment. The inability of E. abietinum to attack the new needles before the autumn (see Development and Physiology section under Biology and Ecology), means that trees always carry some foliage after spring or early summer infestation, or soon regain foliage after autumn or winter attack. This is usually enough to ensure that the trees survive and recover. The death of large or small trees may occur under exceptional conditions, e.g. after complete defoliation in autumn and winter when the trees have been left bare for several months, or when spring attack coincides with a severe frost that kills the new growth, but in Europe this is a rare phenomenon (Bevan, 1966; Straw et al., 1998a). The risk of mortality is higher for small, not fully established trees on difficult sites and when defoliation is followed by drought stress.In North America and New Zealand, defoliation by E. abietinum has caused significant tree mortality. The large number of Picea englemannii and Picea pungens trees killed by E. abietinum in the interior south-west USA is likely to have a significant economic impact, although as yet this impact has not been quantified. Damage by E. abietinum has been considered to be a major factor preventing spruce being used as a production species in New Zealand, limiting options for commercial forestry and potential yields (Miller and Knowles, 1989; Nicol et al., 1998).E. abietinum can defoliate spruce over considerable areas. Carter (1972, 1977) indicated that 37,700 ha of P. sitchensis were defoliated by E. abietinum in western parts of Britain in 1971, which was equivalent to 15% of the total area planted with P. sitchensis at the time. Koot (1991) reported that more than 5000 ha of P. sitchensis were defoliated by E. abietinum on the west coast of Canada in 1981, and Lynch (2002) estimated that E. abietinum defoliated 57,000 ha of P. englemannii and P. pungens in Arizona in 1999-2000. Large areas of P. sitchensis and Picea abies have been defoliated in Europe in other years (e.g. in 1957), but the actual area defoliated or number of trees affected were not estimated at the time (Carter and Halldórsson, 1998). Variation in defoliation by E. abietinum has been related to differences in local climate and type of planting (Ohnesorge, 1961), elevation (Joly, 1961) and soils (Leroy and Malphettes, 1969).E. abietinum has pest status in the production of P. abies Christmas trees in Europe, and in the growing of spruce trees for ornamental purposes, particularly blue spruce (P. pungens var. glauca). Seed orchards of P. glauca in British Columbia are also frequently defoliated (Harding et al., 1998). Aphid control in these situations will also carry a significant economic cost.

Defoliation by E. abietinum, in common with defoliation by other insect pests, influences short- and long-term nutrient cycling in forest stands (Pedersen, 1992; Pedersen and Bille-Hansen, 1999; Stadler et al., 2001). Nutrient availability may be increased in the short-term, but over longer periods defoliation may result in a net loss of nutrients lost from the stand (Pedersen, 1992).Tree mortality in relatively dry areas such as Arizona is of concern because it increases the amount of dry fuel material in the forest and increases the risk of wildfire (Lynch, 2004). This may become an important issue during prolonged periods of drought.

In the interior south-west USA, E. abietinum is a new invasive pest in native, high elevation spruce forests and, on its own or in combination with other factors, is causing the death of large numbers of Picea englemannii and Picea pungens. Complete defoliation of spruce over wide areas has led to 28-42% mortality of more severely defoliated trees and almost total mortality of regenerating seedlings and young saplings (Lynch, 2002, 2004). P. englemannii is more severely defoliated and suffers greater mortality than P. pungens. The mortality caused by E. abietinum in these forests is changing natural disturbance regimes and tree population dynamics, and in the long-term, will cause a shift in species composition of the mixed-conifer forests toward Abies and other non-spruce species (Lynch, 2004). Extensive mortality of spruce in the more isolated mountain ranges of Arizona and New Mexico also poses a threat to rare and endemic elements of the native fauna, e.g. certain species and subspecies of squirrels, which are dependent on spruce in these relict conifer forests.

Old growth stands of P. sitchensis in coastal areas of Queen Charlotte Islands, British Colombia, have a high landscape and amenity value, and the death of such stands following defoliation by E. abietinum has an impact on public use of forest land (Pillsbury, 1960).

In Europe, the first signs that aphids are present are most frequently seen in early spring, although they may also be seen as early as September when the trees are dormant. Individual needles here and there will be yellow or show a yellow band across them where the aphid has been feeding. The best way to see the aphid is to turn the shoot over and examine the underside of the yellowing needles. Placing a sheet of plain white paper under the foliage and rapping the branches with a stick will usually dislodge some aphids if they are present.