Essays 3


The winter moth belongs to the family Geometridae, so the caterpillar is a "looper" with only two pairs of prolegs. It is found throughout most of Europe, East Siberia, Japan, and was accidentally introduced to Eastern Canada.
In the U. K. the caterpillars are active from March to June (Bevan 1987), they emerge at budburst (Carter 1984) and feed for about six weeks on a wide variety of host plants, so the species is polyphagous although individual larvae are monophagous. In the last instar the larvae lower themselves to the ground where they pupate usually in the first 5 cm of soil (Holliday 1977), in a cocoon of silk, late May is the peak time for pupation to start in England (Feeny 1970). The adults emerge between October and February, the female is wingless and crawls up the trunk of the nearest tree to mate, the male has wings and is a night flier. Eggs are laid any time from October to March, they are laid singly (Carter 1984) in crevices of the bark of the tree on which mating took place, each female lays about 150 eggs (Feeny 1970).
Until quite recently the host plants were mainly deciduous trees, oak being the main food plant, and also many fruit trees, birch and hazel. In the last twenty or thirty years the winter moth has expanded its range of food plants to include Calluna vulgaris (Carter 1984), where the larvae have been found in densities of over 1552 m2 (Picozzi 1981) and even Sitka spruce. There had always been small populations on Calluna especially in Orkney where there are no trees, but now the populations on Calluna and Sitka are so large as to be considered pests.
There does not seem to be much discrimination in selection of host plant or oviposition site by the female (Wint 1983), as the choice of larval food plant depends on the female's pupation site and the direction she takes on emerging, as she just climbs up the nearest tree. So it might be fair to say that the larvae contribute more to host plant choice than the adults, as the larvae can balloon if the host plant is unsuitable. However there is the possibility that year after year each generation of females will climb up the same tree as their mother, and oaks (one of the most common host plants) can be hundreds of years old, so there could have been hundreds of generations of winter moth on the same tree. As there is strong selection pressure to time the hatching of eggs to budburst there might be the same pressure to match hatching to their particular tree. Males have much less of a link with individual trees as they fly. The caterpillar feeds on a variety of leaves but a diet of oak leaves has been found to produce heavier pupae than hazel or blackthorn. When Wint (1983) offered freshly hatched and starved larvae a choice of host plant leaves that included Quercus, Malus, Prunus, Crataegus, Corylus and Fagus, they showed a preference for Quercus and Malus. However when larvae were reared on the six host plant leaves the results showed that mortality was not related to host plant, except for Fagus which had 100% mortality.
Budburst synchronisation and ballooning
In early studies in Wytham Wood near Oxford the key-factor in influencing population size was found to be "winter disappearance", this is loss of eggs and young larvae mainly caused by hatching too early (Feeny 1970), i.e. before budburst. So to maximise its chances of survival and reproductive output the winter moth must time its hatching to coincide with oak budburst or be able to use alternative food plants. It has been found (Hunter 1992) that budburst and tree height have a positive correlation with caterpillar density, 75% of the variation in density is accounted for by these two variables. The main influence on budburst is the weather, but even within species in the same area budburst can vary by as much as two weeks. If the larvae hatch when host tree's buds are not open the caterpillars hang from a silk thread till they are blown to another tree, this usually occurs within 12-24 hours of hatching (Wint 1983). They tend to feed on whatever is available and can survive for up to five days without food (Wint 1983). This method of dispersal works best when trees are fairly close together, but is always a cause of high mortality which increases if the trees are widely spaced. The rounded shape of an oak must diminish the chances of the larvae escaping from the tree unless it is at the very top or at the edge, many larvae must land in another part of the same leafless tree. Holliday (1977) grease banded four trees in an apple orchard to prevent females depositing eggs, when densities of larvae and pupae were surveyed for banded and unbanded trees it was found that both had similar densities, so on apple trees in orchards larval density seems to depend on dispersal, not the amount of eggs laid.
Larval predation and parasitism
Perhaps the best known predators are great and blue tits. They time the hatching of their chicks to coincide with the early larval stages, and though each pair may take hundreds of caterpillars to feed their young each day, it has been estimated that this amounts to only 2-5% of the available winter moth caterpillars (Feeny 1970), this may be an example of predator overload. On Orkney larval predators included starling, meadow pipits and common gulls (Picozzi 1981). In Nova Scotia and British Columbia the winter moth was accidentally introduced, to control the populations and protect the fruit crop two parasitiods were imported, Cyzenis albicans, a fly, and Agrypon flaveolatum, a wasp. About five years after the introduction of the parasitoids the winter moth populations declined. About fifteen years later populations in some of the original orchards, now neglected, were examined by Pearsall and Walde (1994). Pupal densities varied widely from over 300 m2 to only 7 m2, and the highest mortality was no longer caused by parasitoids but predation on and in the ground. The parasitoid C. albicans was causing mortality of 0%-20%, and A. flaveolatum was not found at all. So the introduced parasites reduced the populations from epidemic levels to a level similar to that found in countries where the moth occurs naturally, since then generalist predators of the pupal stage have been the main population regulators.
Pupal predation
In Wytham Wood, in England, predation of pupae in the soil mainly by carabids and staphyinids but also by some small mammals, shrews etc. was found to be the main population regulating factor (East 1974). In neglected apple orchards in Nova Scotia Pearsall and Walde (1994) also showed that the population regulating factor was the predation by carabid and staphylinids on pupae. Some beetles could eat 2.5 times their own body weight per day, but beetle numbers showed no relation to pupal density, so it was thought that the beetles just switched to eating winter moth pupae as they became abundant. This introduced population of winter moths has come to resemble the native population (Roland 1994).
The number of eggs laid by the female is related to pupal weight, with heavier females laying more eggs, and the pupal weight is related leaf toughness (Wint 1983) not to the nutritional quality of the leaf. On most trees survival of the larvae is highest if they start feeding at budbust, on oak mature leaves are too tough and both lower weight and increase mortality (Feeny 1970) as the pupal weight is lowered, this decreases the number of eggs laid, and fewer adults survive from eggs laid by females with low pupal weight, so reproductive success is lowered.
Oak is the main food plant of the winter moth at present, but to pupate at a reasonably high weight the larvae have to feed on the leaves as soon as they emerge. Older and mature oak leaves have been shown to be less nutritional because of the increased presence of tannin (Feeny 1970). This increasing tannin content is the cause of spring feeding in the winter moth and possibly in other oak-feeding caterpillars as the tannins complex with the leaf proteins so limit the amount of nitrogen available to the caterpillar (Feeny 1970), like most insects the caterpillars are nitrogen limited.
Oaks that leaf out early, and achieve 50% budburst relatively quickly have the highest density of caterpillars and subsequently suffer the highest levels of defoliation in woodlands. Populations on oak are sometimes heavy enough to completely defoliate the trees. When this happens the larvae must find another food source by ballooning (Feeny 1970). However in parklands where the trees are more widely spaced there is no relationship between budburst and defoliation (Hunter 1992), this may be because the trees are too far apart for the caterpillars on leafless trees to balloon over. In Canadian orchards it has been found that to prevent damage to fruit crops all trees in the orchard and within dispersal distance (perhaps as much as 50 m) must be banded (Holliday 1977). The winter moth has caused defoliation and distorted growth on Sitka spruce recently, this is a new food plant, and as it is grown commercially and planted densely in single species stands there is great potential for population explosions and extensive damage.
True interspecific competition is rarely seen, this is probably because the actual competition has occurred and conditions have settled to an equilibrium, or competition does not occur, or occurs only sporadically. A possible competitor of the winter moth is Tortrix viridana, a leaf-roller. Both caterpillars feed on Q. robur and are on the tree at the same time and at similar life cycle stages. At high densities both are capable of defoliating trees and both have been found within the same bud during their early instars (Hunter and Willmer 1989).
Hunter and Willmer (1989) set up manipulation experiments with single species and two species at different densities from second instar onwards on oak. Both caterpillars showed decreased survival with increasing density, but the winter moth's survival was better when it was with T. viridana than when it was with an equal number of its own species. Pupal weight of the winter moth was higher at a given density in mixed species samples than in single species samples. Which suggests that intraspecific competition may be a greater factor than interspacific competition for the winter moth. T. viridana is a leafroller, this helps it maintain high humidity, when the winter moth's feeding damaged its leaf rolls it suffered higher mortality. So competition can occur between the two species, but only at higher densities or in patches, and the winter moth is the superior competitor.
The winter moth has recently started to feed on Sitka and there have been outbreaks in Calluna, some heathland areas in Holland have been severely damaged. Neither Sitka nor Calluna appear to be very good food plants for the winter moth, yet it is thriving on them despite the high lignin content and thick cuticle. These two species may be more nutritious than they previously were because of increased nitrogen deposition. In Holland conifers are planted in certain areas to uptake volatilised nitrogen from domestic animal waste. In this country NOx from road traffic is a possible source of extra nitrogen. Another reason may be that on spruce mortality of larvae hatching before budburst is not as heavy as that on oak. (Watt and McFarlane 1991). This may be because the scales of buds on Sitka are looser and the caterpillars can enter the bud. This removes the selection pressure to synchronise hatching to budburst which may lead to a division of the species into two races with different selection pressures. Or perhaps the chemoreceptors that tell the caterpillar that Sitka and Calluna are not suitable foodplants have mutated to stop deterring feeding, or to tell it that they are good foodplants.
The winter moth is a very successful insect that, in the past, caused damage to oaks mainly, but with the recent expansion in host ranges to include Calluna and Sitka, both more or less monocultures in their areas the winter moth has the potential to do great commercial damage, especially in a country like Scotland. And the expansion of its range may not have ended yet, it may be able to feed on other conifers.
Bevan, D. 1987. Forest insects,: a guide to insects feeding on trees in Britain. HMSO, London.
Carter, D. J. 1984. Pest lepidoptera of Europe with a special reference to the British Isles. Junk, The Netherlands.
East, R. 1974. Predation on the soil-dwelling stages of the winter moth at Wytham woods, Berkshire. Journal of animal ecology. 43. 611-625.
Feeny, P. 1970. Seasonal change in oak leaf tannins and nutrients as a cause of spring feeding by winter moth caterpillars. Ecology. 51. 565-581.
Holliday, N. J. 1977. Population ecology of winter moth (Operophtera brumata) on apple in relation to larval dispersal and time of budburst. Journal of applied ecology. 14. 803-813.
Hunter, M. D. 1992. A variable insect-plant interaction: the relationship between tree budburst phenology and population levels of insect herbivores among trees. Ecological entomology. 91-95.
Hunter, M. D. and Willmer, P. G. 1989. The potential for interspecific competition between two abundant defoliators on oak: leaf damage and habitat quality. Ecological Entomology. 14. 267-277.
Pearsall, I. A. and Walde, S. J. 1994. Parasitism and predation as agents of mortality of winter moth populations in neglected apple orchards in Nova Scotia. Ecological entomology. 19. 190-198.
Picozzi, N. 1981. Common gull predation of winter moth larvae. Bird Study. 28. 68-69.
Roland, J. 1994. After the decline: what maintains low winter moth density after successful biological control? Journal of animal ecology. 63. 392-398.
Watt, A. D. and McFarlane, A. M. 1991. Winter moth on Sitka spruce: synchrony of egg hatch and budburst, and its effect on larval survival. Ecological entomology. 16. 387-390.
Wint, W. 1983. The role of alternative host-plant species in the life of a polyphagous moth, Operophtera brumata, (Lepidoptera: Geometridae). Journal of animal ecology. 52. 439-450.


Cyanogenesis is the ability of living tissue to produce cyanide compounds. Cyanogenesis takes place mainly in plants and is found widely. There seems to be little pattern in its occurrence except that it may be less common in areas that suffer from prolonged frosts. But how is it possible for a little plant like clover to produce a poison so deadly that it would take only 25 mg to kill a 50 kg human, and why doesn't the plant poison itself?
Scientists have discovered 23 cyanogenic glycosides in plants, and it is believed that there must be many more. They are fairly simple compounds and the ability to make them is controlled genetically. The plants don't poison themselves because they keep the ingredients in separate vesicles, or packages within the cell. One contains the glycoside and the other the degradative enzymes that break it down. When the plant cell is damaged the vesicles rupture, and the two compounds combine to produce poisonous hydrogen cyanide and an aldehyde or ketone. And this is the role that Cyanogenesis is thought to play in plants, it's a defensive compound released to poison animals that try to eat its leaves, they get a mouthful of poison, spit out the leaf and never eat the plant again. At least that is what was thought, but: It doesn't stop ruminants eating clover, they have a bacterial enzyme to detoxify the cyanide.
Many herbivorous insect larvae freely eat cyanogenic leaves and some appear to sequester it and use it as a defence, e.g. the burnet moth. Other insects that prefer acyanogenic leaves can exist perfectly well on cyanogenic ones, e.g. earwigs. The within species variation of cyanogenic to Cyanogenesis individuals bears little relation to herbivory, though there may be some relation to stress, e.g. drought. There is no evidence that plants that have cyanogenic compounds do better than those that do not, just the opposite, on clover acyanogenic plants have more leaves, stolons, flowers and better survival. It's very expensive to produce taking up as much as 25% of the nitrogen in a clover leaf. So it seems that Cyanogenesis does not fulfil the role it was thought to. However it does deter many herbivorous insects from feeding, and if looked at on a population level those acyanogenic plants, because they mimic cyanogenic ones, may also be protected from some herbivorous insects so benefiting the population as a whole.
Other theories are that it is a defensive compound that some herbivores have evolved methods of resistance to, or that it is just a warning that the plant contains nasty chemicals. But perhaps the most interesting idea is that it is not a defensive chemical at all but a means of storing nitrogen for the plant. This would make sense as nitrogen is often in limited supply, plants can resorb unused nectar, and deciduous plants resorb nutrients from leaves before they drop, so perhaps the plant breaks down the cyanogenic glycoside to reuse the nitrogen.

How can we tell whether insect herbivory is genuinely detrimental to trees?

The short answer is that usually we cannot tell for sure that herbivory is detrimental. Herbivory can be classed as genuinely detrimental if it lowers the reproductive success of the plant. Unless the tree dies before reaching maturity it is very difficult to show a lowering of reproductive success, and even more difficult to measure it. But even if you could it is almost impossible to say that herbivory is the cause - it might have been environmental conditions. Also if the tree dies before maturity the cause might not necessarily have been herbivory, again it might be soil conditions, light, etc. But before even trying to pin the blame on herbivory you must show a lowering of reproductive success, and with a long-lived plant like a tree this is not easy.

1. To eliminate bed soil, lack of light, etc. many replicates must be used in identical conditions. This is both difficult and expensive.

2. Reproductive success rates must be measured and known. An oak produces 30 000 acorns a year for over one hundred years. But is that reproductive success? How many of those acorns go on to produce mature oaks? This is the real measure of reproductive success. So you must measure the lifetime output of mature reproducing oaks that were produced from one oak, repeated as many times as possible to ensure replication. Then this must be compared with oaks under the same environmental conditions, but subject to herbivory.

These kinds of experiments are unlikely to be funded these days. So other things are looked at that can be measured in the short term. But the question will always be - Do these things really indicate a lowering of reproductive success?

1. Are there chemical defences produced in response to herbivory? This indicates an allocation of resources that might otherwise have gone to reproduction. It is easy to measure this and the answer is nearly always yes.

2. Do these changes stop or deter further herbivory?

3. Is plant fitness increased as a result of decreased damage?

Chemical analysis of phenolics (these can be seen as plant secondary compounds) for various trees has been done, and showed that the phenolic level varied from year to year, tree to tree, and leaf to leaf, but it did not very with damage to the tree, nor did the leaf biomass vary significantly when the herbivore population was doubled!

There are instances when trees have been completely defoliated and died, but most of these cases were in densely planted monocultures of non-native trees planted on soils less than ideal for the tree, e.g. Sitka spruce on normal to freely draining soil. In such cases the trees suffer stress of one sort or another and may not be able to defend themselves. So the cause of death may be more due to soil conditions than herbivory, and the inadequate soil conditions may have facilitated the insect attack.

So in conclusion we may not be able to tell whether herbivory is genuinely detrimental to trees because it isn't under normal circumstances - it just looks to us humans as if it ought to be.