SPHINGID ECOLOGY


INTRODUCTION

Basic information is provided (from published sources and new field observations) on the general biology, population dynamics, migration, hostplants, and parasitism and predation of the Sphingidae of the western Palaearctic. This is not intended to be an in-depth study but a work aimed at providing a basic understanding of some of the interactive mechanisms which contribute towards the dynamic biogeography of species.

Dynamic biogeography is the study of distribution patterns and biological processes on broad temporal and spatial scales (Hengeveld, 1990). It differs from the study of local spatial patterns and processes, which concentrate on fine-scale phenomena as components of broad-scale phenomena. This latter approach is distribution ecology (Krebs, 1972), where one proceeds from local through regional to global observations.

However, both dynamic biogeography and distribution ecology are not separate or opposing disciplines but represent two approaches to the same phenomena at different scales of variation. Any distinction between biogeography and ecology is simply one of scale over which a phenomenon occurs, combined with the direction in which one looks. At the operational scales of ecologists and biogeographers, the subjects merge into one another -- the distinction between them is artificial and subjective (Hengeveld, 1990).


GENERAL BIOLOGY

The Sphingidae, like all Lepidoptera, are endopterygote insects, i.e. they undergo a complete metamorphosis, passing through four distinct stages -- the ovum (egg), larva (caterpillar), pupa (chrysalis) and imago (adult), (Scoble, 1992).


THE OVUM

The speed of egg development is very variable between species and is not always dependent upon temperature. The eggs of Hyles livornica have been known to hatch in three days; during cold weather those of Sphinx ligustri may take 21 days. Under normal British conditions, those of Laothoe populi populi take 11--17 days to hatch (Williams, 1966). No Palaearctic sphingids undergo diapause in this stage.

Few western Palaearctic sphingids intentionally lay their eggs on anything but their larval hostplants. However, it has been observed that Hyles vespertilio will occasionally oviposit on stones at the base of Epilobium plants, and Rethera komarovi and Macroglossum stellatarum may lay on dead stems protruding through the main hostplant, Galium (Pittaway, 1993). Guided by visual, olfactory, tactile and other stimuli, the female moth normally chooses the spot with great care. However, to quote Reavey & Gaston (1991), 'Where the female moth lays eggs depends on how likely she is to be eaten by predators in the process, how far away she is from plants with nectar for her food, how urgently she wants to lay and where the egg is most likely to survive'. Between one and three ova are usually laid at any one time (up to ten in the Hyles euphorbiae complex), several females sometimes selecting the same shoot. Apart from those species which feed on Galium, where eggs may be placed on the delicate flower-heads, ova are normally deposited on the underside of a leaf where they are hidden and better protected from parasitoids, extremes of temperature and desiccation, or even from being dislodged.


THE LARVA

Unlike birds, young caterpillars eat their way out of the eggshell. After a short rest most consume the remaining chorion and then proceed to a suitable resting-site. No western Palaearctic species is dependent on its eggshell for its first meal. Even the larva of Marumba quercus, which in its first instar does not feed on leaves at all, often ignores the vacated chorion and wanders off.

In some species, the young larva is inactive and sits sphinx-like along the midrib on the underside of a leaf, hidden from predators. In Hemaris fuciformis, small holes are chewed through the leaf on either side of the midrib (Meerman, 1987). In Marumba quercus and Sphinx ligustri, feeding is always from the leaf margin inwards. Many of the Macroglossini prefer the flowers of their hostplants and sit among them, at least while still small. As they grow larger some become strictly nocturnal (Hyles vespertilio passes the day on the ground), or, like Deilephila elpenor and Deilephila porcellus, hide in the tangled mass of hostplant when not feeding. Many of these intermittent feeders change colour from green to a cryptic brown so as to blend in with the dead leaves, or even stones, amongst which they rest. Others, such as the Euphorbia-feeding Hyles species -- Hyles euphorbiae, Hyles tithymali, Hyles dahlii and Hyles nicaea -- feed quite openly and seem to rely on brilliant aposematic coloration to warn vertebrate predators that they are poisonous. Some larvae bear eyespots (Deilephila elpenor, Deilephila porcellus, Daphnis nerii and Rethera komarovi) which may be located so as to make them appear snake-like, thus deterring attack.

Interestingly, mature larvae of many of the smaller Macroglossinae which remain feeding on their hostplants (Hemaris croatica, Sphingonaepiopsis gorgoniades and Macroglossum stellatarum) are cryptically patterned in some shade of green with longitudinal white lines and resemble the younger stages of those which later change colour (Rethera komarovi, Proserpinus proserpina and Hyles vespertilio).

Fully-grown larvae of the Sphinginae practice a different strategy and tend to rely totally on cryptic coloration and patterns to blend in with their hostplant. Most hang suspended under a leaf and move only to feed or relocate. So as not to have their position given away by denuded twigs, larvae of many species (e.g. Laothoe populi) change position frequently.

The growth rate of all caterpillars is very variable and highly dependent on temperature as well as the quality and quantity of food. Little work has been carried out on the nutritional quality of the hostplants of the Palaearctic Sphingidae; Marumba quercus has a higher growth- and survival-rate on Turkey oak (Quercus cerris) than on English oak (Quercus robur) (Pittaway, 1993). Detailed research on the North American tiger swallowtail Papilio (Pterourus) glaucus (L.) (Papilionidae) (Scriber, Lintereur & Evans, 1982) suggests that this would be a fertile field for further study.

Lack of food is rarely a problem for larvae unless too many females select the same small hostplant on which to oviposit, in which case the resultant starving larvae generally wander off in search of new plants before they are mature. Many succeed, some fall victim to predators and a few, if sufficiently developed, will pupate prematurely. However, Hyles livornica, being a true desert species, has developed some remarkable survival techniques to overcome this problem. Although ova are laid on preferred hostplants (Rumex, Asphodelus and Zygophyllum), the resulting larvae can complete their development on a large number of other neighbouring plants should drought or overgrazing eliminate their normal hosts. In fact, under such conditions many larvae actively shorten their instars and pupate when still very undersized (Pittaway, 1993). The resulting dwarf adults are, however, still perfectly formed and fertile.

Unlike those of many other lepidopterous families, no Palaearctic sphingid larvae undergo diapause, although those of Acherontia atropos will become torpid during short, cold spells in the North African winter.

Like all insects, the Sphingidae are poikilothermic and are incapable of maintaining for long an internal body-temperature much higher than their surroundings. As the biochemical processes necessary for growth are faster and more efficient between 20°C and 30°C, both the adults and larvae of the Palaearctic Sphingidae need either to generate or to acquire heat for many activities. At a controlled temperature of 22°C, the larvae of Hyles hippophaes mature in 11--18 days; those of Marumba quercus in 30 days. In central Europe, under field conditions, these species take, on average, 35 and 42 days respectively (Pittaway, 1993). Usually, newly emerged larvae try to avoid intense sunlight but, during cold weather or in northern latitudes, they will actively sunbathe. This activity becomes very noticeable in caterpillars, such as in those of more mature Macroglossum stellatarum, Hyles euphorbiae, Hyles hippophaes and Hyles gallii and allows them to complete their development more rapidly. Large caterpillars can raise their body temperature to 10°C or more above that of their surroundings by this method (Reavey & Gaston, 1991). This ability to condense their life cycles with increasing temperatures and sunshine-rates allows many species to undergo multiple generations in southern Europe and North Africa. It is also possible that Hemaris croatica, Hyles euphorbiae, or even Hyles hippophaes, could not survive so far north in Europe if their larvae did not sunbathe; the amount of available sunlight at critical times of the year may be the main factor limiting the distribution of these species. In Lebanese populations of Hyles euphorbiae, this process has been taken one stage further in that larvae occurring at high altitudes have more black coloration than those from lower, hotter zones and hence can absorb more heat. The same applies to a form of the larva of Hyles nicaea castissima found in the Atlas Mountains (Pittaway, 1993). This black coloration may also protect against high levels of ultra-violet radiation found at altitude.

Once mature, the caterpillar ceases feeding and rests for up to 24 hours. Its colour darkens -- green species generally turn a shade of brown or purple -- and it voids any remaining gut contents. Many tree-feeding species will then descend under cover of darkness and make off rapidly in search of a suitable pupation site.


THE PUPA

At the onset of the pupal stage of metamorphosis, activity is reduced and feeding and excretion cease. Pupation may occur in a subterranean chamber (Acherontia atropos), under stones (Hyles vespertilio), or within a loose cocoon on the surface of the soil (Hemaris, Hyles and Deilephila spp.), or even within leaves on the hostplant (summer broods of Theretra alecto). Larvae of Mimas tiliae and Laothoe populi wander no further than the base of 'their tree' and pupate just beneath surface debris. Occasionally, pupae of Mimas tiliae may even be found high in elm trees (Ulmus spp.) under loose bark. This is probably a method of reducing mortality, for if a caterpillar spends too long and travels too far searching for a safe place to pupate it risks being eaten or losing body weight (Reavey & Gaston, 1991).

During this stage there is a very active reorganisation and transformation of the larval organs; larval tissues are broken down and, in some cases, immediately rebuilt into those of the adult. However, in those pupae undergoing diapause (the overwintering phase for most species) the tissues pass the winter as clusters of nuclear cells in a sea of liquid. (An exception is the pupa of Proserpinus proserpina, which overwinters in an almost fully-formed state.) Only with the onset of warmer weather, coupled with other appropriate stimuli, are they transformed into the adult insect.


THE ADULT

As soon as the internal transformation of the pupa is complete, the adult is ready to emerge. The chrysalis splits dorsally and laterally and the adult crawls out, propelled by waves of abdominal contractions. If underground, these same movements enable the insect to reach the surface. In Deilephila elpenor, abdominal hooks on the pupa allow it to wriggle out of its cocoon and free of leaf debris, before the moth emerges. However, these hooks appear not to have evolved for this purpose but as an aid to escape submergence in water (Brock, 1990).

Once free, the adult moth climbs to some suitable support from which to hang and then expands its wings with haemolymph, a process which takes approximately 30 minutes, after which they dry and harden. Most species emerge in the late morning or early afternoon so that their wings have sufficient time to dry before nightfall.

Prior to their first flight most hawkmoths expel the remaining meconium -- waste matter accumulated during the pupal stage. Apart from the diurnal Hemaris species and Macroglossum stellatarum, and also Hyles tithymali, Hyles gallii and Hyles livornica which occasionally fly in bright sunshine, all western Palaearctic sphingids are crepuscular or nocturnal. None is active the entire day or night and many show definite sexual and feeding behaviour patterns. Some do not feed, but most are avid visitors to flowers. Many are powerful and skilful fliers (Heinig, 1982) and undertake prodigious migrations.

Little work has been done in the field on adult lifespan but in captivity most European species live from 10--30 days; Williams (1966) recorded 5--11 days for females of Laothoe populi populi. The exception is Macroglossum stellatarum, which hibernates as an adult from about October to April.

Before flight, adults must increase the temperature of their flight-muscles. This is achieved by simultaneously contracting both the upbeat and downbeat wing-muscles, which causes the wings to vibrate. The amount of heat generated can be positively correlated with wing area and body size and is independent of ambient temperatures. However, the latter have a marked influence on how long a moth needs to warm up. For example, the North American hawkmoth Manduca sexta (L.) takes twelve minutes to warm up at 15°C but only 1 minute at 30°C (Sbordoni & Forestiero, 1985). At 5°C, this species will not fly at all, although the European Smerinthus ocellatus will (Pittaway, 1993). However, the Sphingidae are not as cold-blooded as many believe. Temperatures within the thorax during flight can exceed 40°C (Casey, 1976; Herrera, 1992); heat-loss from the thoracic muscles is reduced by a dense covering of insulating, fur-like scales. Conversely, other thermoregulatory mechanisms prevent overheating during high ambient air temperatures (Herrera, 1992).

Another method of acquiring body heat is adopted by adults of the diurnal genus Hemaris and Macroglossum stellatarum. Individuals may elect to warm up as normal or, alternatively, bask in the sun. The latter is more energy efficient and it is not uncommon to observe several Macroglossum stellatarum sunbathing on a house wall after hibernation (Pittaway, 1993). In many respects these diurnal species behave like butterflies in response to ambient temperatures and weather conditions.

In larvae, feeding is straightforward and concerned with the efficient acquisition and utilization of food, resulting in growth. In the adult, growth no longer occurs but food is still required to supply energy for activity, to allow ova within the abdomen to mature, and to provide moisture.

Flower nectar is the main source of energy and moisture, but provides little in the way of protein and vitamins. The latter are supplied by the breakdown of body tissues accumulated during the larval stage. Superficially, most adult moths visit flowers indiscriminately, showing little preference for certain colours or species. Careful observations have shown, however, that a moth will not only learn the location of suitable nectar-bearing flowers within its territory, but will also visit each patch at the same time each day or night (Pittaway, 1993). Additionally, many species exhibit distinct periods of feeding activity. In Sphinx pinastri, this is between dusk and midnight, when it avidly visits pale, deep-throated, tubular flowers such as Lonicera. Hemaris tityus prefers blue or purple, shallow-throated flowers such as Ajuga, Salvia and Knautia, which it frequents between 10.00--15.00 h.

Acherontia atropos and Acherontia styx are unique in that they rarely visit flowers, preferring to enter beehives in search of honey. Their proboscises are modified into short, dagger-like tubes capable of piercing honeycombs.

However, feeding and egg maturation take up valuable time and can expose the female moth to increased predation with the subsequent loss of all reproductive potential. The Smerinthini have solved this problem by emerging with large fat reserves and fully-developed eggs. They have no need to feed and can initiate egg-laying immediately after pairing. In consequence, many have atrophied probosci. The drawback of this lifestyle is that few Smerinthini live more than 20 days as adults.

Unlike butterflies, moths are mainly nocturnal and cannot rely on visual stimuli to find a mate. Virgin females of all Palaearctic Sphingidae (even the diurnal species) produce powerful airborne sex pheromones capable of 'calling' males of their own species from some distance. The male approaches upwind, alights and immediately copulates, although in the Macroglossini and Sphingini he may first douse the female in a pungent male pheromone from body hair-pencils (Birch, Poppy & Baker, 1990). Laothoe populi remains paired for up to twenty hours; Hyles euphorbiae for rarely more than two hours. In the diurnal Macroglossinae, the pairing process has developed one stage further -- the sexes may engage in pursuit courtship-flights before copulation and may even dispense with the calling phase altogether (Pittaway, 1993).


POPULATION DYNAMICS

The distribution of hawkmoth species over the western Palaearctic is neither temporally nor spatially evenly distributed. To avoid competition, each species occupies a particular ecological niche comprising an array of fluctuating environmental factors with which it continually interacts and which are utilized by no other species in that combination. These environmental factors are an amalgam of temperature, rainfall, sunshine, seasons, availability of habitat, presence of hostplants and/or nectar sources, and numbers of predators/parasitoids, amongst others (Krebs, 1972). Combined with the genetic parameters of a species, these have a marked effect on abundance and distribution. Grossmueller & Lederhouse (1987) demonstrated that for Lepidoptera in which the adults feed, nectar availability, not the distribution of larval hostplants, controlled local distribution.

Although the climate of the western Palaearctic has four distinct seasons, these do vary from year to year and a given species may experience marked fluctuations in both emergence times and abundance. Many species even stagger emergence to prevent short spells of inclement weather from destroying whole populations. A sphingid species will emerge when:

Although the local population dynamics of a given sphingid can depend on such factors as predation and parasitism rates, migration, and success in mating and oviposition, by far the most important factor is the weather (Krebs, 1972), especially on the periphery of a species range (Hengeveld, 1990).

Spatial variation in habitat quality is probably the main reason why regional populations seldom die out. Within a given area there tend to be certain optimum localities where conditions are favourable every year but colonies produce varying numbers of adults from year to year. Other areas may contain poor localities which, in bad seasons, will fail to produce any adults and are recolonized from adjacent colonies only in subsequent good years.

Thus, in northern Europe most resident species such as Mimas tiliae emerge in May or June. Farther south many of these fly as early as April and, with sufficient warmth, are able to complete several generations in a season. In southern Austria, adults of Deilephila porcellus are on the wing in April and August; in Britain, in June only. Smerinthus ocellatus can produce up to three generations in North Africa but only one in Britain.

A warm spring may encourage a species to emerge early. On 6th May 1986, females of Hemaris tityus were seen ovipositing on Knautia arvensis in the Vienna Woods of north-east Austria after a warm April; a cool spring in 1987 delayed this activity by 20 days (Pittaway, 1993). Such delays may produce only a single generation in a normally double-brooded species in critical areas, whereas hot weather may encourage the development of an extra brood. Poor weather during an entire season may even prevent many larvae from the first or second brood from completing their development. During 1986, in London, southern England, most of the eggs of a partial second brood of Laothoe populi were deposited on Populus spp. (poplar) during late July, but a cool summer coupled with early leaf-fall resulted in significant numbers of immature larvae meeting an untimely end (Pittaway, 1993); the same occurred during 1994 in Oxfordshire. Conversely, the warm summers of 1989 and 1993 allowed this species to complete two full generations successfully in Oxfordshire. Whilst a species can usually recover during later, more favourable seasons, a succession of poor summers or even a slight but gradual deterioration of climate can locally exterminate a species occurring at the limit of its range. This appears to have happened to Hyles euphorbiae in the Netherlands after 1950. During the same period Laothoe populi, a species which is not so dependent on warm, sunny summers, increased in numbers, probably due to the widespread planting of Populus species as amenity trees. A similar study of Mimas tiliae shows a steady decline during the present century (Meerman, 1987). A small but steady deterioration in climate during this century may also have brought about the extinction of Hemaris croatica in central Europe and european Russia, although human factors appear to have been more important.

The number of annual generations a hawkmoth species can produce is directly related to temperature. Usually, the warmer the climate the faster the development. There is, however, another influencing factor -- rainfall and/or moisture (Krebs, 1972). Rarely a problem in northern Europe, a lack of rain in sunnier regions may shrivel hostplants and terminate all breeding activity. In North Africa, Hyles livornica first emerges in the Sahara during October and breeds until May. Thereafter this species either aestivates or migrates north to take advantage of the European summer. The late summer and autumn broods of Hyles tithymali deserticola are also seriously reduced by hostplant desiccation in North Africa (de Freina, 1994).

In some sphingid species, parasitoids are equally as important as the weather in determining seasonal numbers. One colony of Smerinthus ocellatus in London produced 25 healthy and two parasitized larvae in 1985 whereas, in 1986, only two larvae completed their development out of a total of six. In some years upward of 80 per cent of all the larvae of this species in London may be parasitized (Pittaway, 1993). This level of parasitism may have brought about a pattern of later emergence, enabling some sphingids to miss many larval parasites. Sphinx ligustri usually emerges in southern England during mid-June, even though suitable larval hostplants are available from the end of April. However, larvae of this species are prone to parasitism by ichneumonid wasps, which overwinter as adults. By emerging so late it is possible that many potential parasitoids may have already found other hosts and died before suitable larvae of Sphinx ligustri are available.

A combination of good weather and few parasitoids can, in some years, result in a population explosion, which may lead to mass migrations and the colonization of new areas. Annual migrations can have an important influence on the numbers of some species recorded in certain areas. Many of the Sphinx ligustri caught in south-east England each year are migrants from continental Europe. All adults of Agrius convolvuli caught in the Netherlands and the British Isles are either migrants or the first-generation offspring of migrants.

Thus a species may undergo seasonal and/or periodical cycles of abundance and scarcity, and/or show a steady local, area or even regional decline/increase in numbers in response to changing environmental conditions and migration levels. Density-dependent factors (predation, parasitism, disease) and density-independent factors (climate, weather, fires) together help regulate hawkmoth populations, preventing them from either exploding in numbers to plague proportions, or from totally dying out. The dynamic process of this is discussed in detail by Hengeveld (1990).


MIGRATION

Many sphingids undertake 'migrations', but this term has come to mean many things. Is the single moth which leaves a colony on a long-distance flight and is discovered far outside its range, a migrant? Such individuals are best termed 'vagrants' -- opportunistic travellers seeking to found new colonies or join old established ones. Indeed, much of the gene flow between isolated populations of the same species occurs in this manner.

True insect migration is best defined as the directional movement/dispersal of a species, usually massed, from one area to another. However, unlike birds, few Lepidoptera make a return journey to their point of origin -- their lifespan is too short. A species builds up numbers in an area and departs in response to certain, as yet poorly understood, environmental conditions. Thousands of individuals of both sexes can be involved. Most are sexually immature and develop only on reaching their destination. How this destination is determined is unknown, but swarms of migrating moths often pass apparently suitable breeding areas and may eventually alight at a very unfavourable site. For instance, every year small numbers of Agrius convolvuli arrive in Britain, but are unable to establish themselves as they cannot survive the winter here (the author was handed two full-grown larvae during September 1991, which produced moths during October). Such individuals probably originate in West Africa, from where mass migrations to the north have been well studied (Gatter & Gatter, 1990).

How and why migration began is open to conjecture. Species which evolved in hot, periodically dry semi-desert and steppe conditions may have developed migration systems either to take advantage of hostplant growth produced by localized rainstorms or avoid decimation if rains failed in that area in a given year. In this case, the populations become nomadic, shifting location every few years. Hyles livornica has adopted this system, as have many desert butterflies in North Africa and the Middle East (Pittaway, 1981; Larsen, 1987).

A species may also be in an expansive phase, continually pushing at the parameters of its range in order to colonize new territory. The rate at which this occurs depends on the dispersive powers of a species (Krebs, 1972). This may explain the frequent presence of small numbers of typical Oriental species in Palaearctic areas of Afghanistan (Ebert, 1969). This has distinct evolutionary advantages. The last ice age rendered much of the Palaearctic region uninhabitable, forcing many species into Far Eastern and European refuges, as in the case of Sphinx ligustri and Deilephila elpenor. As the post-glaciation climate ameliorated, both species rapidly expanded their range once again.

With migration, a subtropical or tropical species can take advantage of unexploited niches in cooler temperate regions during the more favourable summer months. Macroglossum stellatarum appears to practice this in Europe and many individuals, the descendants of earlier migrants north, have been observed travelling south in autumn. Indeed, in winter, individuals may even migrate south from the Mediterranean and turn up in tropical West Africa (Rougeot, 1972). Whatever the reasons, many western Palaearctic sphingids migrate well beyond their normal ranges. Some, such as Agrius convolvuli, Acherontia atropos and Macroglossum stellatarum, have even occurred as far north as Iceland, where no hawkmoths are endemic (Wolff, 1971).

Vagrants, whether self-propelled, wind-assisted or man-assisted (by boat, plane or accidental release) have been found throughout Europe. Hyles hippophaes has occurred at least once in Britain (Barrett, 1895), north-west Spain (Gomez Bustillo & Fernandez-Rubio, 1985), and Hungary. Some exotics, such as the American Agrius cingulatus and the Asian Daphnis hypothous, probably visit the western Palaearctic regularly, only to be confused with their resident counterparts.

The following exotic species have been captured within the western Palaearctic region, although some have almost certainly originated from breeders and cannot be considered as vagrants:


SPECIESORIGIN
Agrius cingulatus (Fabricius, 1775)North America*
Manduca sexta (Linnaeus, 1763)North America
Manduca quinquemaculata (Haworth, 1803)North America
Ceratomia undulosa (Walker, 1856)North America
Sphinx drupiferarum Smith, 1797North America
Polyptychus trisecta (Aurivillius, 1901)West Africa
Ambulyx lahora (Butler, 1875)North-West India, Pakistan
Pachysphinx modesta (Harris, 1839)North America
Enyo lugubris (Linnaeus, 1771)North America
Eumorpha anchemola (Cramer, [1780])South America
Daphnis hypothous (Cramer, 1780)South Asia*
Nephele didyma (Fabricius, 1775)South-East Asia
Macroglossum nycteris nycteris (Kollar, 1848)North India, Pakistan
Hippotion osiris (Dalman, 1832)Africa*
Hippotion eson (Cramer, 1779)Africa
Theretra boisduvalii (Bugnion, 1839)South Asia*
Theretra japonica (Boisduval, 1869)East Asia

Several of these species, which appear to be genuine vagrants and which are indicated by an asterisk (*), are dealt with in the separate species accounts.


HOSTPLANTS

All sphingids are primary consumers. The adults (with some few non-feeding exceptions) imbibe flower nectar, although some tropical species will drink eye-secretions (Baenziger, 1988) and some North American species have been seen probing decaying animal remains (Sbordoni & Forestiero, 1985). Larvae consume plant material -- they are herbivores.

The most conspicuous herbivores in the region are mammals such as cattle, sheep, goats, rabbits and deer; however, by far the most important terrestrial plant-eaters are the insects. They have evolved a remarkably diverse variety of mouthparts to deal with their hostplants, eating leaves from both the outside and the inside, boring through stems and roots and devouring flowers, seeds and fruits. Considering the abundance of insects and their variety and appetites one wonders how plants still manage to flourish on earth. One reason is the production of defence structures, such as spines, thorns and spiky leaves. Remove these and many insects, or their larvae, will devour the resulting defenceless plant even though they would not normally eat it, e.g. larvae of Sphinx ligustri readily accept leaves of Ilex aquifolium (holly) once the spines have been removed. Another defence is nutritional deficiency. However, all these mechanisms are secondary to the plant world's main line of chemical defence (Pittaway, 1983a).


CHEMISTRY OF PLANTS

Plants produce a wide range of secondary chemical compounds (allelochemicals) in vulnerable parts, which appear to perform no physiological function in the plants themselves but can act as potent repellents, growth retardants or even toxins to insects (Baldwin, 1988). Most widely known is pyrethrin, a Tanacetum (formerly included in Chrysanthemum (Brummitt, 1992; Mabberley, 1987)) extract, which is relatively harmless to mammals but a potent insecticide.

One very large group comprises terpenes, or essential oils, which are simple aromatic carbon-compounds. Most plants contain one or more of these which give them their characteristic smell when crushed, such as menthol in mints (Mentha spp.) (Schauenberg & Paris, 1977).

Alkaloids are equally important. They abound in certain plant-families, such as the Solanaceae, and tend to be more widespread in tropical regions, examples being nicotine, strychnine, caffeine, solanin, quinine, tetrahydrocannabinol and opiates (Schauenberg & Paris, 1977).

Another important group are the acetogenins, such as rotenone and the furanocoumarins. The latter are well represented in the families Rutaceae and Apiaceae (Schauenberg & Paris, 1977).

The Apiaceae are also well endowed with certain types of a very important group of secondary chemicals, the glucosides (heterosides). Depending on their action upon ingestion by mammals, these are classified as flavonoid, cardiac, phenolic, anthraquinone and thiocyanate (mustard) glucosides (Schauenberg & Paris, 1977). Anthraquinones occur in a fairly small number of plant-families, for example, the Rhamnaceae, Polygonaceae and Rubiaceae. Cardiac glycosides are more widespread and are often sequestered by insects for their own defence, such as in the butterfly Danaus plexippus (Linnaeus, 1758). Similar compounds have also been isolated from adults of Daphnis nerii (Chang, 1982) and may perform a similar function.

Related to glucosides are saponins which, together with tannins, raphides (crystals of calcium oxalate), organic acids, and bitter compounds, in addition to those already mentioned, make up the chemical cocktails which protect plants from the attentions of insects (Fraenkel, 1959; Baldwin, 1988; Harborne, 1988; Spencer, 1988).

A gravid female moth intent on depositing eggs is likely to be found in an area where suitable hostplants are present because she herself was bred in that locality. She will home in on likely sites using as yet unidentified stimuli -- probably a combination of chemical, microclimatic and visual. It is her responsibility to select the correct hostplant, that selection being dependent on what the larva can eat. At close range, individual hostplants are located by their chemical 'signature' -- a characteristic mixture of volatile hydrocarbons and other secondary chemicals given off by that plant's metabolism, which are often the very same as those intended to deter insect attack. Before ovipositing, the moth may check a hostplant's identity by means of tarsal receptors, as some completely unrelated plants appear to have very similar signatures. If the signature is not correct, no eggs are deposited and that plant will escape the attentions of that particular insect.

Few if any herbivores (including lepidopterous larvae) have evolved the range of morphological and physiological adaptations which allow them to feed on all plants; most specialize to some degree. Caterpillars employ alimentary mixed-function oxidase enzymes to detoxify secondary plant compounds; the more polyphagous a species the more types and quantities of oxidase enzymes it will generally possess. Bernays & Janzen (1988) did show, however, that different taxa of moths can employ very different methods for dealing with secondary plant chemicals. They studied the Saturniidae and Sphingidae of Costa Rica and found that the large-headed saturniid larvae have short, simple mandibles which cut their food into large pieces of which only the edges are digested. The relatively small-headed sphingid larvae have long, ridged and toothed mandibles which tear and crush food into small pieces. The former generally feed on old, tough, tannin-rich leaves with low levels of secondary toxins, the latter on young, tannin-poor leaves containing high levels of secondary toxins to which they are resistant.

This selection implies that the larva has the ability to distinguish between edible and inedible plants. Caterpillars do bear a range of gustatory sensilla on their maxillae which respond to the variety of secondary and other compounds found in plant tissue (Remorov, 1984). If, during feeding, more positive than negative stimuli are received (all plants contain both beneficial and harmful substances in varying proportions), the plant is edible and feeding either commences or continues. As it is in the insect's interest to have positive stimuli supplied by secondary compounds it is capable of detoxifying, they become feeding- or even oviposition-stimulants for that species. However, if the sensilla are removed, many species can feed on what appear to be unsuitable hostplants (Remorov, 1985).

To protect themselves from the most important terrestrial herbivores, the insects, plants therefore produce a range of chemical compounds which vary, both in proportion and type, between species, genera and families. However, the full role of plant secondary compounds is uncertain as conflicting results have been obtained from laboratory studies (Reavey & Gaston, 1991).

This situation is not static, however, but in a constant state of evolutionary flux for, to avoid being fed upon, a mutant plant which can produce a new deterrent will prosper at the expense of its fellows which cannot. But, as the resistant plant population builds up, a deterrent-factor pressure is put on the original insect population to adapt to this deterrent. Some will be able to respond; some may change to hostplants containing deterrents similar to the original hostplant. Thus, for those insects which 'crack' a new deterrent-code, competition will be reduced and often leads to speciation; this appears to have happened with Hyles chamyla in Central Asia, a species which almost certainly split off from Hyles hippophaes quite recently.

Similarly, were a polyphagous species to shift its range to a new area devoid of its original hostplants it might, through a drift in genotype, lose the ability to feed on those plants. Conversely, if there were no selection pressure against that factor, it might persist and express itself if the moth and its ancestral hostplant were reunited. This appears to have occurred in the European populations of Sphinx ligustri, which are thought to have originated from an ancestor in western North America (Hodges, 1971) and spread to Asia and then Europe. The larvae of this species normally feed on Fraxinus, Ligustrum, Syringa and Spiraea and, occasionally, Lonicera. Some of the related North American species prefer Symphoricarpos, a genus of the family Caprifoliaceae naturally confined to that continent, especially the western portion. However, with the introduction of Symphoricarpos alba into Europe, larvae of Sphinx ligustri are to be found on this plant in ever-increasing numbers (Meerman, 1987), as well as on the introduced Forsythia (from China) and Physocarpus opulifolius (from North America).

Most work on lepidopterous hostplants has been concerned with butterflies, so, in order to understand the present day pabulum of the European hawkmoths, it is worthwhile looking at the Papilio machaon complex of swallowtail butterflies. Papilio machaon Linnaeus is the world's most widely distributed swallowtail, ranging across Europe, North Africa, northern Asia and northern North America (Tyler, 1975). In central and southern North America it is partly replaced by closely related species, such as Papilio indra Reakirt, Papilio polyxenes Fabricius, Papilio brevicauda Saunders, Papilio zelicaon Lucas and Papilio joanae Heitzman. In Europe and North Africa there are three other related species, namely Papilio hospiton Guenée, Papilio alexanor Esper and Papilio saharae Oberthuer. All can feed in the larval stage on the taxonomically unrelated but chemically similar Rutaceae and Apiaceae (Thompson, 1988). Rutaceae contain furanocoumarins, small amounts of tannin, a bitter compound, the glucoside rutin, and the monoterpenes methyl chavicol, anethole and anisic aldehyde. Although rutin may be absent, all the above are present also in the Apiaceae. Differences in the selection of wild hostplants appear to be due mainly to adult behaviour for, although Papilio machaon and Papilio polyxenes will not normally oviposit on Asteraceae containing the same or similar secondary compounds, their larvae can be reared on Tagetes and Cosmos. However, two members of the complex, Papilio machaon bairdii Edwards and Papilio machaon oregonius Edwards, have overcome the as yet unidentified oviposition deterrent and adopted Artemisia dracunculus as their natural hostplant. Additionally, certain populations of Papilio machaon in Afghanistan and Alaska have also transferred to the genus Artemisia.

Hawkmoths show similar preferences in their selection of hostplants. The chemical relationships are as important as the evolutionary relationships between plant species. Indeed, the ability of most sphingid subfamilies and tribes to detoxify only certain secondary compounds is so characteristic that one could almost classify them according to these abilities (Harris, 1972; Pittaway, 1983a).

The orders Myrtales, Euphorbiales, Rhamnales and Rubiales must contain similar secondary compounds to enable the Macroglossinae to feed widely amongst them, or to specialize on one particular family in an order (see list of the principal hostplant families below).


Many botanists include the Elaeagnaceae in the order Myrtales, a view supported by its adoption by Hyles hippophaes for its hostplants.


NUTRITIONAL FACTORS

Why do several female moths oviposit on the same plant when apparently equally satisfactory, if not better, plants of the same species are all around? Why are some species of hostplant preferred to others? Why do larvae of Marumba quercus reared on Quercus cerris thrive better than those on Quercus robur (Pittaway, 1993)? Why do many of the Macroglossini prefer the flowers of their hostplants rather than the leaves? That the biochemical composition of leaves varies, both between and within species, is not in doubt (Reavey & Gaston, 1991). However, although considerable research has been undertaken regarding the chemical interactions between plants and insects, very little is known of the nutritional quality of known hostplants (Schroeder, 1986).

Recently, various American researchers have produced some interesting results from studies of the survival rate of the larvae of Papilio (Pterourus) glaucus (Linnaeus) (tiger swallowtail) on a variety of hostplants under conditions of an average day-length of sixteen hours and a temperature range of 19.5--23.5°C. Not only were differences recorded in the results obtained with different hostplants (Table 1), but further research by Scriber & Lederhouse (1983) demonstrated differences within that one species at different times of the year (Table 2). Papilio glaucus is usually bivoltine, or univoltine with a bimodal emergence.



Table 1: Survival and development potential of larvae of Papilio (Pterourus) glaucus (Linnaeus), studied at Wisconsin, U.S.A. in 1979 (after Scriber, Lintereur & Evans, 1982).

Hostplant% survival to 3rd instarNo. days as larvaPupal wt. (mg dry)
Magnolia acuminata6735.4249.0
Liriodendron tulipifera3630.6301.1
Sassafras albidum4635.6213.2
Lindera benzoin0--
Betula papyrifera5741.8315.6
Populus tremuloides0--
Sorbus americana8041.5272.9
Prunus virginiana7540.2248.7
Prunus serotina5732.4335.4
Ptelea trifoliata0--
Fraxinus americana50--


Table 2: Developmental duration of larvae of Papilio (Pterourus) glaucus (Linnaeus) reared in 1973-74 on two hostplants under field conditions at Ithaca, New York, U.S.A. (after Scriber & Lederhouse, 1983)

HostplantNumber of days as a larvaNumber of days as a larva
June--JulyAug.--Sept.
Prunus serotina36.3 (7810.8*)47.0 (8030.4*)
Fraxinus americana44.5 (9345.1*)55.0 (7973.3*)
* actual accumulated thermal units

Research into the hostplant preferences of the various western Palaearctic sphingids would undoubtedly produce equally interesting results and to a limited extent has already done so for Laothoe populi (Table 3). Grayson & Edmunds (1989a) found that larvae reared on Salix fragilis tended to pass through only four instars, whereas those on Populus alba underwent four or five. However, although more larvae survived on Salix fragilis and fed up more rapidly, the adults from larvae reared on Populus alba were larger and more fecund.



Table 3: Survival and development of larvae of Laothoe populi populi (Linnaeus) on two hostplants, given as the mean of three trials (after Grayson & Edmunds, 1989a)

Hostplant% larvae surviving to 7 days% larvae surviving to 21 days% larvae surviving to pupationNo. of days as larva
Salix fragilis76.053.344.030.2
Populus alba50.023.014.645.3


PARASITES AND PREDATORS

As moths have been on the earth for more than 190 million years, numerous other organisms have evolved to exploit this rich food source.

Very few adult moths are subject to lethal parasitism, probably due to their limited lifespan and inability to grow -- they do not accumulate tissue-mass or food-reserves. Eggs and caterpillars, however, are subjected to severe attacks by parasitoids in addition to the numerous fungi, bacteria, viruses and nematodes (Siebold, 1842; Hagmeier, 1912; Fiedler, 1927; Mathur, 1959) which take their toll. Although nematodes have been found in several species, most notably Hexamermis albicans in Mimas tiliae (Hagmeier, 1912), little is known of the exact effects they have on the host. Parasitoids have been studied in greater detail (Hammond & Smith, 1953, 1955; Askew, 1971; Shaw & Askew, 1976; Herting, 1984; Shaw, 1982, 1990).

Tiny wasps (Trichogrammatidae) may lay their eggs through the micropyle of a hawkmoth egg and develop within. In Iran, many ova of Rethera komarovi manifica are lost in this way (Pittaway, 1993). Indeed the Macroglossini seem particularly prone to this type of parasitism. As the surviving larvae grow, various members of the Tachinidae (Diptera), Ichneumonidae and Braconidae (Hymenoptera) take an interest in them. Belshaw (1994) gives an account of the various ovipositional and behavioural strategies employed by the Tachinidae, and Gould & Bolton (1988) give a good overview of the biology of parasitic Hymenoptera. Most of the tachinids deposit their eggs on the larval cuticle or the leaf on which hosts are feeding. These hatch in a few days and the parasitic larva burrows into the host's body to feed on its tissues. It is often only after the host pupates in the ground that the larval parasitoid emerges through the cuticle and itself pupates. In North Africa numerous larvae of Acherontia atropos fall victim to Drino (Zygobothria) atropivora (Robineau-Desvoidy), whilst Drino vicina (Zetterstedt) and Lydella stabulans (Meigen) take their toll of Hippotion celerio (Rungs, 1981).


Adult of Masicera sphingivora, Germany. Photo: © Tony Pittaway.

The various parasitic Hymenoptera develop somewhat differently. Their eggs are usually laid through the host's skin, although in those few species which develop as ectoparasites, they are placed on the cuticle. In the Braconidae, the mature endoparasitic larvae usually vacate the host in its final instar and spin individual but coalesced cocoons around the host remains, often after it has prepared its own subterranean pupal chamber. Up to 80 per cent of all Smerinthus ocellatus larvae in Britain are lost every year to Microplitis ocellatae Bouché, a gregarious braconid. In China this parasitoid is equally devastating to populations of Smerinthus planus (Wei & Zheng, 1987).


Adult Microplitis ocellatae, England. Photo: © Tony Pittaway.

Whereas many braconid larvae may develop in a single host caterpillar, in the Ichneumonidae one parasitic larva per host is the norm. Most of the species that attack sphingids pupate within the host's pupal skin and emerge through a hole in the cuticle. Many of the Ichneumonidae are host-specific or restricted to two or three genera; the Braconidae display a preference for the Smerinthini in their relations with sphingids.


Adult of Pimpla illecebrator, England. Photo: © Tony Pittaway.

Recent work in the U.S.A. has shown that the hostplant selected by a caterpillar may determine its susceptibility to attack by a given parasitoid (Barbosa et al., 1986). It is possible that this factor plays as great a role in forcing sphingid species to change hostplants as does hostplant resistance.

Being subterranean, no pupa of any western Palaearctic sphingid is known to be directly parasitized, although it is strongly suspected that Protichneumon pisorius (Linnaeus) is capable of digging down to the pupae of Sphinx pinastri and parasitizing them (Hinz, 1983).

Numerous animals, both invertebrate and vertebrate, find hawkmoths and their larvae tasty morsels. Wasps of the genus Vespa and Vespula destroy countless shrub- and tree-feeding larvae every summer, such as those of Sphinx ligustri and Laothoe populi. Mice, shrews, birds, bats, ants, beetles and spiders also take their toll and, were it not for the fact that the Sphingidae have developed a variety of sophisticated defensive and protective strategies (Evans & Schmidt, 1990), many species would become extinct.


DEFENCE

Moths and their larvae are often the victims of predators using visual skills, and many defensive mechanisms have evolved to counteract these. However, bats are almost unique among aerial predators in hunting moths by echo-location and not sight. Adult moths of the families Arctiidae, Noctuidae and Geometridae can detect the ultrasonic emissions of a bat and take avoiding action. This may explain why these moth families are very successful and why, in the Sphingidae, the Choerocampina, with their 'ear' (Roeder, 1975), are one of the most successful sphingid subtribes. However, it is thought most deter or escape capture by their size (many hawkmoths being as large as the insectivorous bats themselves) or by means of their flying skills and speed (Rothschild, 1989; Evans & Schmidt, 1990).

A number of larvae exhibit cryptic coloration, blending in with their hostplant. Their outline can be further broken up by deportment, body stripes and countershading. The larva of Smerinthus kindermannii always rests upside down under a narrow leaf with the anterior of the body raised. To enhance its disguise as a leaf further, it has lateral stripes and a pale dorsal surface which counteracts any shade the body may cast (referred to as 'countershading'). When this position is reversed and it is turned the other way up, this effect is lost and the caterpillar becomes very visible. The effectiveness of this mechanism has been demonstrated by Edmunds & Dewhirst (1994).

To avoid diurnal parasitoids and predators, many sphingid larvae have become nocturnal, hiding on the ground or low down in/on their hostplant by day. In such species the larva changes colour from a cryptic green when young to a mottled brown which blends in with dead leaves and stones. In Hyles vespertilio, such a strategy works well (Milyanovskii, 1959), but it does not prevent the larvae of Deilephila porcellus being actively sought out by tachinid flies. However, the lower the larval density the more chance of escaping detection, so 'spacing out' helps reduce the odds of discovery (Tinbergen et al., 1967; Meerman, 1988a).

Several adult moths also utilize cryptic mimicry, blending in with the vegetation among which they rest (Rothschild, 1989). Laothoe populi resembles a cluster of dead leaves, an illusion further enhanced by hindwings which project above the forewing margin. Mimas tiliae breaks up its outline with various blotches and has a leaf-like wing shape.

If discovered, some hawkmoths can suddenly display brightly coloured hindwing patches to surprise and disorientate would-be predators and enable them to drop to the ground unnoticed (Blest, 1957). Such patches -- usually red, a known warning colour -- remain concealed beneath cryptically-coloured forewings when the moth is at rest. All Hyles species are so endowed.

The three species of Smerinthus found in the western Palaearctic augment the warning effect of the red on their hindwings by bearing false eye-spots. These resemble vertebrate's eyes and, in conjunction with a beak-like abdomen, may imitate the face of a small owl and so intimidate predators. Many inexperienced young birds retreat when so confronted. Similarly, some hawkmoth caterpillars bear striking thoracic eye-spots which can be enlarged by expanding the thorax, as do the larvae of Daphnis nerii, Rethera komarovi, Deilephila elpenor and Hippotion celerio. In these the anterior segments are trunk-like and can be contracted, giving the appearance of a small serpent. Such larvae often lash from side to side like a striking snake.

Many lepidopterous larvae can also assimilate toxins from their hostplants and sequester them in their tissues. Such larvae are often brightly coloured and feed quite openly yet unmolested. Hyles euphorbiae, Hyles dahlii, Hyles tithymali and Hyles nicaea all feed as larvae on species of Euphorbia, and are known to be poisonous to birds -- the main sequestered toxin (a diterpene, ingenol ester) being highly inflammatory and carcinogenic. It is interesting to note, however, that the adult moths contain relatively low concentrations of this carcinogen (Rothschild, 1985). Ingestion of such larvae often induces vomiting in birds as large as a crow. Indeed, one such experience has such a lasting effect that other larvae of that species, or similar-looking species, are shunned. Larvae of Daphnis nerii sequester small quantities of cardiac glycosides in their tissues which are passed on to the adult stage (Chang, 1982). Larvae of Acherontia atropos seem to confine plant toxins to their large gut and so gain only partial protection from them; they rely more on cryptic coloration for protection, as do the larvae of Daphnis nerii.

However, an increase in the uptake of toxins into the body tissues can be achieved by utilizing alternative hostplants. Larvae of the North American Manduca sexta, when fed on Nicotiana (tobacco), excrete the nicotine they ingest. When reared on Atropa belladonna (deadly nightshade), larvae sequester and store atropine and synthesize two metabolites, which are found in the frass. The resulting pupae (and probably adults) are toxic to chickens (Rothschild, 1985). This ability to sequester and store varying levels of toxin depending on the hostplant eaten can also greatly affect the success of parasitoids attacking larvae of a given species (Thorpe & Barbosa, 1986; Beckage et al., 1987, 1988).

Few adults of western Palaearctic Sphingidae inherit the unpalatability of their larvae, apart from the partial protection acquired by the Euphorbia-feeding Hyles species and Daphnis nerii. However, high concentrations of histamine (75 ug/g) have been found in adults of Smerinthus ocellatus, and acetylcholine in the male accessory gland and ejaculatory duct of Laothoe populi (Rothschild, 1985).

Many perfectly edible butterflies and moths mimic poisonous species to gain protection, a feature known as Batesian mimicry. For example, adult Hemaris species mimic bumble bees (Bombus spp.). The latter, with their painful sting, are studiously avoided by most avian predators. In addition, the adults of Hemaris fuciformis are also distasteful, thus exhibiting a degree of Muellerian mimicry, where several equally poisonous species resemble one another for added protection. In this case, predators learn to avoid the warning characteristics -- the toll taken before they do so being distributed among the co-mimics. What type of mimicry is involved in the close resemblance of the larvae of Hyles nicaea to those of Hyles centralasiae is not understood. The former is known to be poisonous to vertebrates; the latter may or may not be, although its hostplants (Liliaceae) are.

When attacked, some hawkmoths are capable of emitting sounds. The larva of Acherontia atropos audibly clicks its jaws together; the adult hops about and squeaks. Some adult male hawkmoths will also stridulate their genitalic valvae and emit volatile pyrazines from abdominal hair-pencils if attacked (Rothschild, 1989). What effect these noises and smells have on potential predators is not known.

Most sphingid larvae will regurgitate their sticky, sometimes toxic, foregut contents over the attacker. This may be accompanied by violent lashing movements from side to side and is extremely effective against parasitoids and ants. Marumba quercus rarely regurgitates but it employs this lashing action to great effect against parasitoids -- tachinid flies can be impaled on the head spines (Pittaway, 1993). If the lashing fails, the caterpillar of this species will aggressively bite its attacker, a habit it shares with the larva of Acherontia atropos.

Finally, many of the larger sphingid adults are armed with formidable tibial spurs which can draw blood if the moth is handled or attacked.


SUMMARY

Dynamic biogeography is the study of distribution patterns and biological processes on broad temporal and spatial scales (Hengeveld, 1990). It pulls together and interprets the results of studies on biodiversity, and local spatial patterns and processes, such as biology and distribution ecology. Such processes affect all four stages of a sphingid life cycle (Krebs, 1972), with some being of more importance to certain life stages than others. Thus before one can understand the biogeography of a species one must fully understand its ecology and biology.

Adult emergence is governed by many biotic and abiotic factors, the most dominant of which is the climate, particularly sunshine levels and temperature (Krebs, 1972). For most of the species of the western Palaearctic, emergence must take place when it is warm enough for adult activity and survival, yet early enough to allow complete larval development, especially in climatically marginal areas. In areas providing sufficient accumulatable thermal units (warmth) for more than one generation per season, these must be so timed so as to take full advantage of the situation without jeopardizing species survival or waste excessive genetic material. Populations of many species, such as Laothoe populi populi, often 'risk' a proportion of their genetic resources in marginal areas and try to squeeze in a partial second or third generation. In some years this is successful, in others it is not and many larvae meet an untimely end with leaf fall. Univoltine species tend to have extended or even bi-modal periods of adult emergence to avoid spells of inclement weather destroying the entire population. A significant proportion of the pupae of each brood of bi- or tri-voltine species go into diapause for similar reasons, especially in species which live in climatically erratic areas, such as the deserts of North Africa. In such species, e.g. Hyles tithymali deserticola and Hyles livornica, a proportion of pupae also often diapause for more than one season (Pittaway, 1993). This is an insurance measure against periodic droughts which may produce extended periods which are unfavourable for larval development. Like many species of desert butterfly (Larsen, 1987), Hyles tithymali deserticola and Hyles livornica even appear to use the periodicity of rainfall to trigger emergence.

Climatically adverse periods, such as cold winters or hot, dry summers, are thus generally avoided by diapausing as a near-inactive, non-feeding pupa. Such a method of diapause must confer significant advantages as only one species, Macroglossum stellatarum, hibernates as an adult, and no species of sphingid diapause as ova.

Once emerged, adult behaviour is governed by three drives -- the need to find food, the need to find a mate and reproduce, and the need to disperse the species. Adult Smerinthini have dispensed with the food-finding stage, which probably allows areas with no sources of nectar to be exploited and reduces adult predation rates as females need not expose themselves to danger while searching for food. The drawbacks of this lifestyle are reflected in a reduced life-span and fecundity compared with the Sphingini and Macroglossinae. Female Smerinthini have to emerge with all their ova fully formed within them as they cannot ingest additional nutrients for egg development. Weight limitations and a reduced life-span probably dictate, respectively, the maximum numbers of eggs a female can carry and still fly, and a physical limit to the number of eggs which can be laid during her life.

Ova tend to be preferentially deposited at sites which reduce female mortality to a minimum during egg-laying (Reavey & Gaston, 1991) and which provide sufficient local sources of nectar for the female, if she feeds (Grossmueller & Lederhouse, 1987). These sites usually maximize egg and larval survivability due to decreased levels of predation, parasitism and unfavourable abiotic factors, such as dislodgement, desiccation and low temperatures. The behaviour and coloration of larvae is aimed mainly at reducing levels of predation and parasitism while trying at the same time to exploit hostplants and climatic parameters to the full. Predator/parasitoid avoidance can take three forms -- cryptic coloration to avoid detection, hiding behaviour and nocturnal feeding, and aposematic coloration/sequestration of toxins. At range margins, the need to obtain sufficient warmth for full development by exposed basking may result in detrimental rates of predation and parasitism. Conversely, larval basking may allow a species to extend its distribution beyond what ambient summer air temperatures would allow. This appears to be the case with Hyles euphorbiae, Hyles h. hippophaes and Hemaris croatica in central Europe. On the other hand, too much heat and too few spells of cool weather during the cold season may physiologically preclude cold-zone adapted species which diapause from colonizing southern regions, such as Mimas tiliae in Turkey. Prolonged low winter temperatures have a similar effect on cold-sensitive thermophilic species, such as Marumba quercus in the Ukraine.

Species continuously push at the boundaries/margins of their ranges (Hengeveld, 1990) by producing vagrants and by migration. Should conditions become more favourable beyond a range margin then a species will 'creep'/disperse forward to establish a new margin, as has been documented by Hengeveld (1990) and Krebs (1972), and for Acherontia styx in the Arabian Peninsula (Wiltshire, 1986; Walker & Pittaway, 1987). True migration overcomes the vicissitudes of a changeable and mosaic environment, such as is found in deserts, and provides the chance of colonizing new zoogeographic areas beyond the normal range of that species, e.g. Agrius cingulatus colonizing the Cape Verde Islands. Few vagrant species penetrate the western Palaearctic from other regions due to formidable barriers at the boundaries; however, some notable migrants do appear in huge numbers from the Afrotropical region, e.g. Agrius convolvuli and Acherontia atropos.

Another major factor affecting sphingid distribution is the distribution and quality of hostplant(s). Adult emergence must be so timed so as to not only provide developing larvae with adequate amounts of hostplant of high nutritional quality, but take place when competition is at a minimum. A sphingid can be excluded from climatically suitable areas by a lack of hosts (Krebs, 1972) or invade new areas when suitable hosts become available, as has happened with Hyles h. hippophaes in the Aegean (Pittaway, 1982a). Yet, the entire range of one host or a range of hosts may not be exploited due to other, unfavourable parameters, such as climate, historical factors, competition or even host quality (Krebs, 1972). Cool summers and a lack of sunshine exclude Hyles h. hippophaes from northwestern Europe even though its hostplant is locally common.

Parasites may fulfil an equally large role in sphingid distribution, but this is as yet poorly understood. They certainly play a major role in local species abundance; however, parasitoids and predators are as equally affected by climatic conditions as are their hosts (Krebs, 1972). A poor season for a parasitoid can result in a good season for the sphingid host.

Just as the various sphingid life stages have numerous behavioural and ecological traits which allow them to counteract the detrimental effects of climate, parasitoids and predators (Krebs, 1972), so too do their hostplants. Plants produce a wide range of allelochemicals in vulnerable parts which act as potent repellents, growth retardants or even toxins to insects. Like most herbivores, sphingid caterpillars have evolved a range of behavioural and physiological adaptations (mainly alimentary mixed-function oxidase enzymes) which allow them to counteract some of these chemical defences and hence feed on selected (usually related) hosts with similar chemical constituents. In effect, these secondary chemicals have become the stimuli by which female moths can identify suitable hostplants and which induce larvae to feed. Such chemicals are even sequestered by some larvae, such as those of Hyles euphorbiae, to render them poisonous to many predators and parasitoids. However, depending on what they can detoxify and/or tolerate, different sphingid species have preferred and secondary hosts upon which growth rates differ. This allows a species to switch hosts should its preferred diet be locally scarce, e.g. Mimas tiliae on Sorbus aucuparia in the Alps. The larvae of most species cannot afford to invest large amounts of energy and resources in producing the complete range of alimentary enzymes required to detoxify the toxins found in all plants. A compromise seems to take place which also takes into account niche specialization to reduce competition.

Thus the ecology and biology of the western Palaearctic Sphingidae must therefore be regarded as a dynamic yet balanced process involving the interplay of many biotic and abiotic factors (Hengeveld, 1990). The unbalancing of any one of these factors, be it by man or natural climatic changes, usually results in the sphingid species affected becoming extinct, decreasing or increasing in range, retreating to a refuge, or exploding in numbers until a new balance is achieved.


Return to species list