SPECIFICITY IN PREDATOR—PREY RELATIONSHIPS

seilcoAL WITH PARILIQULAR REFERENCE TO

THE APHIDOPHAGOUS COCCINELLIDAE

R. L. Blackman B.Sc.

A thesis submitted for the Degree of Doctor of Philosophy of the University of London

Imperial College of Science and Technology, Field Station, Silwood Park, Sunninghill, ASCOT, Berkshire. October, 1965. 2.

ABSTRACT

In the first section of the thesis an attempt is made to present a picture of the nature and extent of specificity as it occurs in the food relations of certain groups of predacious insects, in so far as the present state of our knowledge permits. The experimental work described in the second section of the thesis was done on two species of aphidophagous Coccinellid, Adalia bipunctata L. and Coccinella septem- punctata L. The effects of different aphid species on the rate of development and mortality of the larvae and the size and fecundity of the adults are compared. The relative unsuitability of certain aphids, especially Aphis fabae as food for Adalia bipunctata, is discussed. A comparison is made of the feeding behaviour of the two Coccinellids on four species of aphid. Experiments are described in which the ability of larvae and adults to select a suitable aphid or avoid an unsuitable one was investigated. The ecological significance of the results is considered, and the potential value of such predators in pest control, given a better knowledge of their prey relations and specificity, is briefly discussed. CONTENTS

SECTION 1. INTRODUCTION AND REVIEW OF THE LITERATURE.

INTRODUCTION 6 SPECIFICITY IN PREDATOR-PREY RELATIONSHIPS 10 1. Insects which are predatory in the larval stage only 10 2. Insects which are predatory in the adult stage only 18 3.Insects which are predatory in both feeding stages a) Adults and larvae with different feeding habits 22 b) Adults and larvae with similar feeding habits 27 SPECIFICITY IN THE COCCINELLIDAE 39

SECTION 2. THE SUITABILITY OF DIFFERENT APHIDS AS FOOD FOR ADALIA BIPUNCTATA L. & COCCINELLA 7-PUNCTATA L.

INTRODUCTION 61 REARING TECHNIQUES AND GENERAL PROCEDURE 62 THE EFFECT OF DIFFERENT APHID FOODS ON THE DEVELOPMENT OF THE LARVAL PREDATOR 68

(A) Adalia bipunctata: 1) Rate of Development 68

4.

2)Weight of Food Eaten 72 3)Mortality during Larval Development 77 4)Weight of Adult at Emergence 79 (B) Coccinella 7-punctata 82 THE EFFECTS OF DIFFERENT APHID FOODS ON THE FECUNDITY OF THE ADULT PREDATOR. 87 (A)Adalia bipunctata 87 (B)Coccinella 7-punctata 92 APHIS FABAE AS FOOD FOR ADALIA BIPUNCTATA 97 A COMPARATIVE STUDY OF THE FEEDING BEHAVIOUR OF LARVAE OF A. BIPUNCTATA & C. 7-PUNCTATA ON FOUR APHID SPECIES. 107 SELECTION BEHAVIOUR BY THE LARVAL PREDATOR 120 (A)Adalia bipunctata • 120 (B)Coccinella 7-punctata 131 SELECTION BEHAVIOUR BY THE ADULT PREDATOR 136 EXAMINATION OF THE GUT CONTENTS OF FIELD COLLECTED ADULTS 145

DISCUSSION 150 SUMMARY 161 ACKNOWLEDGEMENTS 165 BIBLIOGRAPHY 166 5.

SECTION 1

INTRODUCTION AND

REVIEW OF THE LITERATURE. 6.

INTRODUCTION

The predatory habit has been developed many times independently in insects. Some members of every major insect order are predacious. As a result predacious insects considered as a group are an extremely diverse assembly, incorporating a wide range of structure, feeding habits and choice of food. The situation is further complicated by the fact that an insect may be predacious in both larval and adult stages, in the larval stage only, or in the adult stage only. Accounts of the occurrence of predatism among insects have been given by Sweetman (1936), Clausen (1940), Balduf (1936,1939) and Brues (1946). It is somewhat surprising that Balduf (1942) is the only writer to attempt any sort of classification of the types of predator-prey relationship occurring in insects. He uses three main divisions; (1) Hyphenic relations, in which he places all 'occasional predators', irrespective of whether they meet the rest of their food requirements as phytophags, scavengers, or parasites; (2) Absolute predatism, characterised by the fact that 7.

the food taken by one or both feeding stages "consists exclusively of insect prey", and (3) Opportunistic feeders, insects which accept any food that comes their way, exemplified by the feeding habits of most Tric ioptera. As an example of an 'occasional predator' in group (1), Balduf cites the Anthomyid fly Hylemya cilicrura Rond, of which the larva, although a major pest of various crops throughout the world, is essentially a general feeder and has been recorded feeding on living and dead animal matter, particularly the eggs of locusts. He also includes in this group Lycaenid caterpillars which feed on plants in the earlier instars but complete their development by devouring the brood in the subterranean nests of ants. The change of diet is here brought about by the production from the second instar onwards of an ant- attractirv; secretion by the caterpillar, as a result of which it is tolerated in and may even be carried into the ant-nest. The change of diet is thus the result of a special adaptation on the part of the Lycaenid, which becomes an obligatory predator for the rest of its larval life. 8.,

Thus Balduf's 'hyphenic relations' may be obligatory, or facultative on the part of the predator entering into them. It would seem more logical to make a basic distinction between facultative and obligatory predation. Facultative predators may resort to predation in certain circumstances, perhaps even more often than not, but adoption of the predatory habit is not essential to them. Obligatory predators do not necessarily feed exclusively by predation, and in fact many predators use other food sources when their prey becomes scarce, but the predatory habit is essential to their survival and development at one or more stages in the life-cycle. Further consideration of the types of facultative predation occurring in insects would be outside the scope of this thesis. The discussion of specificity in prey relationships of obligatory predators depends on an understanding of the different types of relationship occurring between predators and their prey. It is important to distinguish between insects in which only the larvae are predatory, insects in which only the adult stage is predatory, and insects which are predatory in both feeding stages. In the 9.

last of these groups, which is much the largest, it is necessary to make a distinction between insects where the larvae and adults are markedly dissimilar in their feeding habits, and insects where both feeding stages have similar food and habitat. The following consideration of specificity in predator-prey relation- ships is based upon this scheme. Only the more important groups of terrestrial insect predators will be considered, and as mentioned already discussion will be restricted to obligatory predators. 10.

SPECIFICITY IN PREDATOR-PREY RELATIQNSAIP

(1) Insects which are predatory in the larval stage only.

Balduf (1938, 1939) has given a very complete account of the occurrence and development of the predatory habit in the Lepidoptera, and this forms a convenient and interesting starting point for the review. Entomophagy in larvae of Lepidoptera was first reviewed by Clark (1926). Eight families of Lepidoptera contain members with predacious larvae, yet with very few exceptions in all these families Coccids are the only prey. Balduf's theories as to the origin of entomo- phagous habits in Lepidoptera attempt to explain this almost exclusive coccidophagy. In the Lycaenidae, larvae of some species eat lichens and algae on the bark of trees, whereas other species in genera of the same sub-family prey on scale-insects in the same sort of situation. A change of feeding habit from eating lichens to eating Coccids is not difficult to envisage, and Balduf proposes that such a change has come about many times in different families of Lepidoptera. Feniseca tarquinius Fabr., the first caterpillar known to be entomophagous (Edwards 1886), is an exception to the general rule of coccidophagy. This species feeds on several species of aphids, but all are woolly aphids, sessile insects similar to Coccids in that they would not involve a significant change in feeding behaviour from a lichen-feeding habit. Other minority foods of predacious caterpillars include ant larvae(Liphyra Lycaena spp., Cyclotorna) and Jassid and Membracid guests in ants' nests (Megalopalpus, Lachnocnema); both these habits having arisen according to Balduf's theories from the entrance of ants into the lives of the caterpillars. The ants either build the caterpillar into their nest incidental to sheltering the Coccids on which it is feeding, or guide or transport the caterpillar to their nest because it yields an attractive secretion. In the latter case the Lepidopteran may be deprived of its original food and resort to other food-sources, such as the ant-brood and ant-guests. The egg-laying habits of the adults of coccido- phagous Lepidoptera are imperfectly known, but in general the eggs seem to be deposited in close association 12.

with the larval food, and are often inserted in between or laid on the backs of individual Coccids. The range of species preyed upon clearly depends on the degree of specificity shown by the adult in its oviposition behaviour. There is some evidence of specificity, in the relationship between Eublemma amabilis Moore and lac scales (Laccifer spp.) for example (Misra, 'Negi and Gupta, 1930), but other species show little sign of any preference in their choice of Coccid as larval food, and prey records are too inadequate for any definite conclusions. It seems possible that the adult butterfly is attracted by the Coccid honeydew, and this is another way in which the caterpillar- Coccid association could have arisen. Two other groups in which larvae only are predacious will be considered. These are members of two families of Diptera, the Syrphidae and the Cecidomyidae. Syrphid larvae are best-known as predators, although the family as a whole shows a considerable range of habit; many species are phytophagous, some are aquatic and saprophagous, one genus is myrme- cophilous and another lives as a scavenger in the nests 13.

of wasps and bumblebees. The main food of predacious Syrphid larvae, with a few inevitable exceptions, is aphids. One species with an exceptionally wide range of prey is Xanthan.drus comtus Harris, which although generally aphidophagous is also a common predator of various caterpillars including the pine processionary Cnethocampa pinivora Tr. (H.D. Smith 1936). Syrphid larvae, apodous and without special optic organs or proper biting mouthparts, seem ill-fitted for the predacious habit, and soft-bodied, comparatively helpless insects would appear to be the only prey generally available to them. They are largely dependent for their food supply on the oviposition site chosen by the adult. The oviposition behaviour of adult Syrphidae has been the subject of some disagreement (T.J. Dixon, 1959). The most generally accepted view seems to be that expressed recently by Bombosch (1962), that aphidophagous Syrphidae lay their eggs close to aphid colonies, and only very rarely on uninfested plants. Dixon (1959) has shown that alfactory stimuli enable Syrphus luniger Meigen to locate and oviposit close to aphids. Recent work by Chandler (unpublished) suggests, 14.

however, that there has been excessive emphasis on the influence of the aphid stimulus due to the concentration by previous workers on one or two Syrphus spp.; other Syrphids such as Platycheirus peltatus Meigen, generally those which lay eggs in batches rather than singly, oviposit in relation to a suitable aphid host plant but not necessarily in the presence of aphids. Such a dichotomy would mean that ovipositing Syrphids can exhibit specificity in two different ways, either by choice of host food plant, probably using visual stimuli, or by choice of aphid, in which olfactory stimuli undoubtedly play a part. Another aspect of specificity concerns the food preferences of Syrphid larvae and the suitability of different aphid species as food. The relative importance of this depends on the degree of selection shown by the adult in its choice of oviposition site, but it is nevertheless deserving of serious study. Ninomiya (1957) has presented results of feeding tests involving 12 species of Syrphidae fed on 40 species of aphid, which appear to show a considerable variation in the degree of choice shown by different Syrphid larvae. No studies on the effects of different aphid 15.

foods on larval development have so far been undertaken. The predacious larvae of Cecidomyidae are still more limited in their powers of lodomotion and feeding adaptations than Syrphid larvae, and are accordingly restricted to certain prey. Even so, as a group they have a comparatively wide food range, which is reviewed by Barnes (1929, 1930 1933). The most common prey are aphids, Coccids and mites, the individual genera of Cecidomyids being restricted to one group or another for their food. There has been a surprising tendency to regard Cecidomyid larvae as generally monophagous (Clausen 1940, p. 354), even to the extent of naming newly-discovered species after the prey on which they were found. Observations by Nijveldt (1954) and Milne (1960) on Phenobremia spp. indicate that the range of prey may in many cases be wider than was previously thought. Very little work has been done on the feeding habits of Cecidomyid larvae, and none at all on the oviposition behaviour of the adults; but in many cases elucidation of taxonomic relationships is necessary before such work can have real value. Thus, in the Lepidoptera and in the Dipteran 16.

families Syrphidae and Cecidomyidae, the predatory habit has been developed in the larval feeding stage only, from phytophagous or saprophagous origins. The change of habit does not appear to have resulted in any marked changes in feeding behaviour or morphological structure, the predacious larvae of these groups depending on certain food sources to which they are restricted by limitations imposed by their ancestry. Probably there has also been little accompanying physiological change. The gregarious, relatively sessile, phytophagous Homoptera are the main source of food in all three cases. The predacious larvae of Lepidoptera depend primarily on Coccids, whereas Syrphid larvae, lacking proper biting mouthparts, are restricted to soft-bodied insects and feed predominantly on aphids. Predacious Cecidomyid larvae as a group utilise a wider range of prey; although also with reduced mouthparts they can attack Coccids perhaps because their small size enables them to penetrate under the scales. Individual species in these three groups are generally restricted to one type of prey; there are very few instances where the food range is wide, but 17.

also absolute monophagy occurs rarely if at all. The degree of specificity io primarily determined by the oviposition behaviour of the adult. In order to ensure a food supply for the larvae, the fecund female requires a behaviour pattern which will orient her to the vicinity of the larval food. Since the adult insect has a different feeding habit, the pattern of behaviour leading to oviposition and the reactions to visual or olfactory stimuli involved in it will be an act set apart in the life of the adult insect. In Lepidoptera with phytophagous larvae selection of a specific host plant by the ovipositing adult is commonplace, and it would be surprising if some butterflies with predacious larvae did not also show a marked degree of specificity. In aphidophagous Syrphidae, reaction to olfactory stimuli does play a part in oviposition behaviour, and provided that some aphids have specific odours then there is the possibility that a high degree of specificity has in some cases been developed. In this context, taking an example from outside the three groups just considered, the observations 18.

of Parker (1918) on the predacious larva of the Ohloropid Chloruisca_glabra Meigen are of interest. This is one of the few predators recorded in this Dipteran family. The larvae feed on the sugar-beet root aphid Pemphigus betae Doane, and have been recorded only on this species. The eggs are deposited in the soil at the base of the plant, and the female flies show remarkable ability to select their oviposition site. Eggs are deposited in numbers only around plants which are infested with root aphids. Oviposition is preceded by a period of aimless wandering on the plant, during which the fly presumably determines the presence or absence of root aphids.

(2) Insects which are predatory in the adult stage only.

Relatively few insects have developed the predatory habit in the adult stage only. Some sawflies (Hymenoptera, Tenthredinoidea) prey on small Diptera and Coleoptera on flower-heads (Benson 1950), but whether they are obligatory predators or not is unknown. Records of the prey taken by British 19

Tenthredinidae are given by Hobby (1932); the predatory individuals are mainly females, and they appear to be unseledtive in their choice of prey. Other sawflies, including males of their own species, may also be attacked. In the Diptera many families have developed the predatory habit in the adult stage, and in some it appears to be obligatory. In the same families the larvae also are sometimes predacious, but such instances are comparatively rare and many of the records are unsubstantiated. For simplicity, only the more important families of Diptera in which the adults are predominantly predacious, but the larvae are not or have not been proved to be so, are included in this section. The two families of predacious flies most adequately recorded are the Asilidae and the Empididae. An early but very useful review of prey records for these two families is given by Poulton (1906). The Asilidae appear to have a very wide range of prey, encompassing eight orders of insects. Most common prey are Hymenoptera, of which the majority are bees 20,,

and wasps. Powell and Stage (1962) record observations on prey selection by Stenopogon engelhardti Bromley, and conclude that this Asilid takes a wide range of prey, preference being oriented towards easy to capture insects (size range 4-18 mm.), and controlled by the availability of species. There food was plentiful, there was selection for slower-flying, more conspicuous types. Empidid flies, however; prey predominantly on other Diptera. Although some other insects are invariably taken, the preference for Dipteran prey seems to be remarkably consistent in all species of the family for which there are adequate records. There is a considerable size range within the family; larger species such as Empis opaca Meigen capture relatively large prey (Hobby and Smith 1961), smaller species take smaller insects (eg. Hilara maura F.; K.G.V. Smith 1952), but the prey is mainly Diptera in all cases. It is worth noting that in many Empidinae epigamic behaviour by the male involves catching prey and offering it to the female for her to suck during copulation. It woad be interesting if the preference for Dipteran prey was connected with this behaviour. 2L,

Many other families of Diptera have members recorded as predacious, but information is mostly inadequate for general conclusions. Adult Dolichopodidae, for instance, are thought to be wholly predacious, but records of the species of prey taken are very rare indeed. In the higher Diptera, the dungflies (Cordyluridae) appear to be characteristically predacious, and Hobby (1931a) records the prey of Scopeuma (= Scatophaga) stercorarium L. as being predominantly Diptera. Some Anthomyid flies are predacious, and Hobby (1934) has noted adaptations to the predatory habit in the mouth- parts of some species (eg. Coenosia tigrina P., Prosalpia sylvestris Fall.) The prey taken by Coenosia spp. is again mainly Liptera, Anthomyidae predominating (Hobby 1931b). The predatory habit has also been developed in the adults of some species of Calliphoridae. The specialised food habits of the genus Bengalia are described by Bequeart (1922)c Several species subsist by robbing ants which are in the act of transporting their larvae, pupae, or insect prey, while another 22.

species captures and feeds on winged termites. Thus, in the sawflies and in many Dipteran families, adult insects have developed a predatory habit which is independent of and bears no relation to the feeding habits of their larvae. They have utilised a food source which is not available to larval forms, i.e. other free-flying insects of the highest and most active orders of insects, In certain Diptera, especially the Empididae, there is evidence of a food preference for members of their own order. The Diptera are probably the most

abundant group of free-flying insects, however, •• and this must be taken into account in assessing any degree of specificity. No instances of extreme specificity have been recorded, except where specialised feeding habits are involved, such as in species of Bengalia.

(3) Insects which are predatory in both feeding stages.

(a) Adults and larvae with different feeding. habits.

Instances where both larvae and adults of a species are predacious but with different food and habitat occur in many families of Diptera and Coleoptera, but 23..

such scattered occurrences contribute little in the present context and will not be considered further. Pit-making is a larval feeding adaptation which has been developed in more than one insect order and warrants some consideration. A detailed account of the known feeding habits of ant-lions (Neuroptera; Myrmeleonidae) is given by Balduf (1939). Ants are the principal food but not the only food of the larvae; the pits are only found in situations where ants are common, and worker ants are obviously the most likely insects to fall into such traps. The adult insects are said to lay their eggs in the soil. The degree of correlation between the oviposition site and the presence of ants' nests does not appear to have been investigated. There are also very few records of the food of adult Myrmeleonidae. In the Diptera Rhagionidae, the genera Vermileo and Lampromyia have developed similar habits in the larval stage to the ant-lions, in spite of a very diff- erent bodily structure. Again the prey is reported to consist mainly of ants. Adult Rhagionidae are said by Clausen (1940) to be consistently predacious, the prey consisting mainly of adults of other Diptera, but it 24.

is doubtful whether this is definitely established. It is not kLowsi whether the eggs are laid in relation to ants' nests. A larval feeding adaptation of a similar kind is that found in the tiger beetles (Carabidae: Cicindelinae), where the larva extends the hole made by the ovipositor of the parent beetle to form a burrow which it occupies throughout larval life. The prey appears to consist of any small animal which comes within reach of the larva as it lies at the entrance to the burrow, including centipedes, spiders and Isopods as well as all kinds of insects (Shelford 1909). In the Cicindelinae the site occupied by the larva throughout its life is wholly determined by the oviposition site chosen by the adult; but there is no record of adult beetles laying in relation to a particular source of food. Records of the prey of adult tiger beetles indicate that they take an even greater variety of food than their larvae, including worms, small crabs and any available living animals which they can overcome. Thus adult and larval Cicindelinae are alike in the food which they take but differ in their manner of obtaining it. 25.

At this point it is convenient to consider the larval food-provisioning habits of fossorial wasps and ants. As regards specificity the fossorial wasps are very interesting; the range of prey preferences of the group as a whole is wide, including all the main insect orders and the spiders. Considering each family separately, however, the range of prey is considerably restricted; the Pompilidae provision their nests with spiders (see Evans and Yoshimoto 1962), the Eumenidae mainly with Lepidopterous larvae, the Ampulicidae with cockroaches, the Crabronidae mainly with Deptera, and so on. Individual species are generally still further restricted in their choice of prey. For example, most of the European species of Crabronidae specialise in one family of Diptera (Hamm and Richards 1926); one species, Anothyreus lapponicus Dahlb. is recorded as taking only the Rhagionid Symphoromyia crassicornis Pz. There are other instances of specificity in the fossorial wasps; for instance, the Sphecid, Sphex lobatus F. on the cricket, Brachytrupes portentosus Licht. (Hingston 1925, 1926), and the Philanthid, Aphilanthrops frigidus Smith which prov- isions its nest only with queen ants of the genus Formica, 26

captured during the nuptial flight (Wheeler 1913). Pre-imaginal conditioning has been suggested to account for the remarkable powers of prey recognition of fossorial wasps, and in particular to explain the existince within some species of local races differing in their food preferences (Hamm and Richards 1926). A detailed analysis of the numerous observations and prey records published in recent years (eg. Bristowe 1948, Krombein 1961, arandi 1959 a, b) would provide a better basis for an understanding of specificity in this interestingeroup. In contrast to the fossorial wasps, all records of the prey of carnivorous ants indicate that the Formicidae generally lack any kind of prey specificity. The driver and legionary ants of the tropics are notorious for the way in which they destroy all forms of animal life when on the march. The wide range of food provided by Lasius niger L. and Lasius flavus F. for their larvae is recorded by Pontin (1961); as well as most orders of insects it includes earthworms, woodlice, Phalangids end pseudo-scorpions. The only examples of specificity are found in the Pomerinas; according to 27.

Wheeler (1926) the Texan Lobopelta elongata feeds mainly if not exclusively on Isopoda, and certain species of Leptogenvs, Lobopelta, Ophthalmopone and Diacamma prey mainly on termites.

(b) Adults and Larvae with similar feeding habits.

Similarity of feeding habits between adult and larva is found quite generally in the Hemiptera, and in the Coleoptera and Neuroptera, orders of holometabolous insects in which the divergence between immature forms has not reached the same level as in the Lepidoptera, Diptera and Hymenoptera. Again the lack of information about many groups with predatory habits makes it nec- essary to limit consideration to only a part of the complete field. The food and feeding habits of the Neuroptera have been reviewed by Killington (1936) and by Balduf (1939). Discussion here will be confined to the Hemerobiidae, the Chrysopidae and the Coniopterygidae, the three families which have been most investigated, and are also of most economic importance. Larvae of all Neuroptera are carnivorous, and have mouthparts adapted for piercing and sucking. Larvae 28;

of the terrestrial species are predacious, mainly on small Hemiptera (Tingidae, Aphididae, Coccidae, Aleyrodidae, Jassidae and Psyllidae), Thysanoptera and Psocoptera, but have also been observed to attack mites, small spiders, insect eggs, and small larvae of Coleoptera (Coccinellidae and Phytophaga), Diptera (Syrphidae) and Lepidoptera (Pteridae, Tortricidae, etc.). Although aphids are undoubtedly more preyed upon than other groups, larval Neuroptera are clearly relatively unspecific in their choice of food. In general, larvae kept in captivity will feed on any relatively inactive animal of suitable size which they encounter. It is also possible that some Chrysopid larvae, eg. Chrysopa carnea Stephens (Rabaud 1926), feed frequently on nectar. The absence of specificity in the larvae, however, is often compensated by the habitat and oviposition site selected by the adult. In some species of Hemerobiidae and Coniopterygidae, the larvae are confined to certain trees or types of tree, and are consequently restricted in the food which is available to them. For example, Sympherobius pygmaeus Ramb. appears to be confined solely to oaks, and Coniopteryx pygmaea End. has only been found on conifers (Withycombe, 1923). Of the 12 British species of the genus Hemerobius, six are confined in the larval stage to conifers (individual species showing a preference for larch, spruce or Scots pine) and six are found almost exclusively on deciduous trees and shrubs (Killington 1936). Larvae of the Chrysopidae also tend to be restricted to certain habitats, and may be roughly divided into aboreal species, low vegetation species, and an inter- mediate group that is more widely distributed. This classification is suggested by Balduf (1939) on the basis of the published records of the distribution of the adults. He says that it is 'usually applicable' also to the larvae. Records for the larvae are far less numerous, but there are indications that they fit such a classi- fication even better than the adult. Taking the British species of Chrvsopa as an example, larvae of C. Playa, vittata, ciliata, flavifrons, ventralis and dorsalis are woodland species, the last named being confined to conifers; C. phyllocroma is found always on low vegetation in open fields, and C. abbreviata is always associated with vegetation on coastal sand-dunes, such 30.

as marram grass; C. albolineata, carnea, septempunctata and perla belong to the intermediate group and are more widespread in their distribution (Killington 1936). Adult Neuroptora take the same wide range of prey as their larvae. Examination of the gut contents of Hemerobiidae and Chrysopidae by Killington (1932) showed that the food was mainly aphids, but other prey included, in several species, small adult Lepidoptera. Various species of Hemerobiidae and Chrysopidae were supplied with Microlepidoptera in captivity, but were only observed to feed on injured or dead specimens, and Killington concluded that adult Lepidoptera which serve as food under natural conditions must be caught when resting or injured, or even consumed when dead. Thus, in the Neuroptera, specificity appears to be based primarily on habitat selection. Many species are associated with only certain types of vegetation, and hence are restricted in their range of prey species, but there seem to be no actual food preferences. In those species with a relatively wide range of habitat the range of prey is correspondingly wide. In captivity, larvae which are confined to conifers under natural 31.

conditions will take readily to a diet of aphids from deciduous trees or herbs. Members of several families of terrestrial Heteroptera have developed the predatory habit, although probably relatively few are obligatory predators. The families Reduviidae, Nabidae, Phymatidae and Anthocoridae are generally considered to be predominantly predacious. The Mirid Cyrtorhinus mundulus Bredd. provides an important example of host food plant specificity in a predator. C. mundulus is an efficient egg predator on the sugar-cane leaf-hopper Perkinsiella saccharacida Kirk. In captivity, it will live and breed upon maize, and feed upon the eggs of the corn leaf-hopper, Peregrinus maidis Ashm., as well as upon Perkinsiella eggs on sugar-cane. Muir (1931), however, found that under natural conditions Cyrtorhinus mundulus did not occur on maize even when it was growing among sugar- cane and was heavily infested with corn leaf-hopper, while the surrounding sugar-cane contained very few sugar-cane leaf-hoppers yet had numbers of Cyrtorhinus. Information concerning the food and feeding habits of predacious Hemiptera is chiefly limited to a few species which have been used in biological control, and a generalised discussion is impossible. Further consideration of this group will be restricted to a brief review of recent work on Anthocoridae. Anderson (1962a) records the preferred habitats and prey of six species of Anthocoris in England. Anthocoris nemorum L. is ubiquitous and polyphagous, but the other five species all show some degree of selection in their choice of habitat. A. minki Dohrn. is restricted to ash, A. sarothamni Douglas and Scott tends to be confined to broom, A. gallarum-ulmi DeG. is found predominantly on elm, A. nemoralis F. and A. confusus Reut. tend to be restricted to certain deciduous trees. Anderson found that the life-cycle of A. minki is closely correlated with the occurrence of the larvae of the Psyllid, Psyllopsis fraxinicola Foerster, the predominant insect on ash in Southern England. The larvae of A. minki emerge while the Psyllids are still early instar larvae, and both the prey and the predator complete larval development by mid-July. Likewise the life-cycle of A. gallarum-ulmi which breeds in the 33 .

curled leaf-galls formed by Eriosoma ulmi L. on elm, is closely correlated with that of the aphid. These two Anthocorids appear to have only one generation a year. A. sarothamni, however, has two generations a year on broom, and its life-cycle was found to be associated with the occurrence of the larval stages of two Psyllids, Psylla spartiophila Foerster and Arytaina genistae Latr., although the broom aphid Acyrthosiphon spartii Koch was also preyed upon. Hill (1961) found that in Scotland the life-cycle of A. sarothamni coincided with that of Acyrthosiphon spartii, and he believed that the aphid and Arytaina genistae are the most important prey species there. A. nemoralis appears to breed primarily on Psyllid- infested trees, and A. confusus is found predominantly on aphid-infested trees, beech, oak, lime and sycamore. A. nemorum shows no distinct host-plant preferences, except for a tendency to aggregate on Salix in early spring, and seems to be a general predator. Hill (1957) records it as preying on 35 species of insects and mites representing 10 orders. Anderson (1962b) also investigated the development 34*

and fecundity of the six Anthocorid species on different prey under laboratory conditions. Psyllids proved to be very suitable prey for the larvae of all species, and were significantly better than aphids for A. sarothamni,

...•1••••••••••••A. minki and A. nemoralis. All species, however, could be reared on other than their natural prey, and A. gallarum-ulmi which is monophagous under natural condi- tions was successfully reared on three other aphids and one Psyllid. Anderson's conclusions regarding the food of the three Anthocorids which occur on broom at Silwood Park were substantiated by Dempster (1963), using the precipitin test. He was able to show that on broom A. nemoralis and A. sarothamni feed selectively on Psyllids, but there was no evidence of any selection by A. nemorum. These six Anthocorid species present an interesting picture. All but one (A. nemorum) show a distinct specificity. In A. nemoralis and A. confusus specificity appears to be oriented towards the prey, the former feeding mainly on Psyllids and the latter on aphids. Both species breed on a relatively wide range of host 35..

food plants. In A. minki, A. gallarum-ulmi and A. sarothamni, specificity may have different founda- tions, in the selection and recognition of a host food plant, development of a definite association between the predator and a specific host food plant leading to a closer relationship between predator and prey. Further work on host food plant recognition and oviposition site preference are needed to establish this last point. Carabidae are well-known predators, yet, if we exclude the Cicindelinae already discussed, the number of species in the family which can be regarded as obligatory predators may be quite small. Lists of feeding records are given by Balduf (1935) and by Davies (1959). Forbes (1880, 1883) examined the gut contents of 178 individuals from 38 species of Carabidae in North America, and concluded that taking the family as a whole, at least half the food was plant material. Lindroth (1949), on the basis of feeding studies on 138 Scandinavian species, concluded that Carabidae were "as a rule omnivorous". Davies (1953) studied the gut contents of British Carabids, dissecting 366 specimens of 84 species. His results although agreeing 36s

essentially with those of Forbes showed certainly fairly clear-cut food preferences, and help to clarify the picture, at least as far as the adult beetles are concerned. The sub-family Carabinae appears to be predomin- antly carnivorous, and certain specialisations of feeding habit occur. The members of the tribes Carabini and Cychrini have mandibles without molar processes and practise extra-oral digestion. These Carabids therefore tend to specialise in large soft-bodied prey which they can pierce and suck. The Carabini have a wide food range, including earthworms, snails and even nestling birds, but larvae of Lepidoptera are the most frequently recorded food. Lack of specificity in food selection is apparent in Calasoma frigidum Kby., which accepted the larvae of one or more species belonging to eighteen different genera of Lepidoptera (Balduf 1935). In the tribe Cychrini, Donisthorpe (1913) found both adults and larvae of Cychrus caraboides L. feeding on snails. The Notiophilini are characterised by their large eyes and speed of movement. Davies examined the gut contents of six species of the genus Notiophilus and 37..

found that Collembola and mites were the predominant food in all of them. A few pollen grains and fungal spores were also present but these were thought to be the secondary food of the Collembola. He concluded that Notiophilus is adapted to the specialised role of predator on Collembola and mites in the leaf litter of the woodland floor. The gut contents of Leistus spinibarbis F. (Nebriini) and Loricera pilicornis F. (Loricerini) revealed in these species also a tendency to specialise on Collembola. One example of a high degree of specialisation is found in the genus Dyschirius (tribe Scaratini). Both adults and larvae of most species of this genus feed exclusively on the larvae and pupae of fossorial Staphylinids of the genus Bledius. Their habitat is restricted to particular soil conditions on sandy coasts and on the borders of inland waters, and Lindroth (1949) suggests that their dependence on Bledius is because the Staphylinid is the only constant inhabitant of this habitat fit for prey. The sub-family Harpalinae shows a much greater variety of feeding habit, and it is possible that it 38.

contains no obligatory predators. The tribes Zabrini, Amarini, and Harpalini are essentially plant-feeding, and their mandibles are adapted for this purpose. Forbes (1883) found, however, that the Harpalini take some animal food. The Feroniini, and possibly also the Bembidiini, appear to be true scavengers (Davies 1953). Other tribes such as the Lebiini, however, may be essentially carnivorous. Food records of Carabid larvae are rather limited. The larvae are generally quite active, and their feeding habits where known appear to resemble those of the adults. 39.

SPECIFICITY IN THE COCCINELLIDAE.

In the preceding pages of this review an attempt has been made to present a picture of the extent and nature of specificity as it occurs in the food relations of certain groups of predacious insects. A comprehensive review of this subject is impossible in the present state of our knowledge, and many groups about which insufficient is known have been omitted. Discussion of isolated examples of specific and non-specific predators has also been avoided, for this would not give any general indication of the extent to which specificity or polyphagy occurs in the groups of which they are members. The main criterion for inclusion or exclusion of a group from the review is the extent of our knowledge concerning the group as a whole. Only by this means can a valid picture of specificity in predator-prey relationships be presented. The more that is known about a group of predators, the more coherent becomes the pattern of food relations which it exhibits. This becomes very apparent when we examine the prey relationships of the predacious 40.

Coccinellidae. The Coccinellidae are the best known of all insect predators and also the most adequately recorded. In 1928, Schilder and Schilder produced an extensive critical review of all published records of the food of the family up to that date. Further feeding records are listed by Balduf (1935), but these do not significantly affect the general conclusions reached by Schilder and Schilder, which will be summarised here. For the present purpose the sub-family Epilachninae can be dismissed as phytophagous. The Coccinellinae are, however, almost entirely carnivorous, and it is likely that the majority are obligatory predators in both larval and adult stages. Exceptions are the tribe Psylloborini, which are fungus-feeders, and species of the genus Verania (Veraniini), which are phytophagous, specialising on Graminae. In Coccinellinae a high degree of specialisation exists at the tribal level. Homoptera are the predominant prey of most tribes, but the Hyperaspini feed mainly on Coccids (75% of the references), and 74% of the records of Coccid food refer to the Eriococcid 41.

genera Pseudococcus, Phenacoccus and Ripersia. The Microweiseini are also predominantly Coccid-eating (75% of the reports), but prefer the Diaspididae (Aspidiotus, Chionaspis, etc.); this family comprises 73% of all the references to Coccid food. Coccids are again the predominant food (62%) of the Scymnini, but in this tribe the individual genera tend to specialise in different directions. Thus Stethorus preys exclusively on Acarina (reviewed by Kapur 1948), Clitostethus and Lioscynus prefer Aleyrodidae. Several genera specialise on Pseudococcus and its close relatives; these include Cryptolaemus, for which 92% of the reports of Coccid food are of Pseudococcus spp. on a wide variety of plants all over the world. Pullus feeds on aphids, especially Phylloxera, and the large genus Scymnus is again mainly Coccid-eating. The Pharini and Telsimiini feed on Coccids of the family Diaspididae, whereas the Azyini appear to prefer Lecaniid scales. In the Exoplectrini, the genera Rodolia and Novus feed almost exclusively on Icerya and its close relatives (94% of all reports). The Scymnillini 42.

and the Oneini prey entirely on Aleyrodidae. The Cheilomenini and the Ceelophorini are mainly aphid predators. The large tribe Chilicorini feed mainly on Coccids (79% of the reports). Aspidimerus, Strictobura and Orcus prey exclusively on Coccids, but avoid the genus Pseudococcus. Only Brumus, which will occasionally take other prey besides Coccids, and possibly also Curinus, are recorded as feeding on Pseudococcus. The European and North American species of Exochomus live mainly on Coccids (71% of the reports), yet it is interesting that the African species feed on aphids. Chilicorus is again a Coccid-feeder which avoids Pseudococcus, except for the three most common European and North American species which also feed on aphids. Two remaining large tribes feed on aphids. Seventy-six per cent. of the records for Hippodamini are of aphid food, the remaining records being of Coleoptera (9%); Heteroptera (5%); Lepidoptera (3%); Coccidae (2%); Acarina, Thysanoptera and plant material (each 1%); Diptera, Cicadidae and Fulgoridae (each i%) The references to Coleoptera, Heteroptera and Lepidoptera 43.

as food chiefly apply to the genera Hippodamia and Coleomegilla, which occasionally take the eggs and larvae of these groups. In the Coccinellini aphid food is even more predominant (85% of the reports). The remaining records are of Lepidopteran eggs/larvae and Coccidae (each 3%), Coleopteran eggs and larvae (2i%), Acarina and Psyllidae (each 2%), Heteroptera and plant material (each 1%), Thysanoptera and Hymenoptera (each 1-%). The results of the work of Schilder and Schilder have been presented fairly fully in order to give an idea of the types of specialisation in feeding habit found in the Coccinellidae as a whole, and the different systematic levels at which these specialisations occur. It is interesting to consider the probable phylogeny of the family (Strouhal 1927). The Epilachninae are regarded as the primitive Coccinellidae, from which the Coccinellinae are derived. Members of the tribes Coccinellini and Chilicorini are perhaps closest to primitive Coccinellinae; from these tribes were derived highly specialised groups such as the fungus-eating Psylloborini and the Eriococcid-eating Hyperaspini 44.

respectively. The Scymnini are a tribe with both unspecialised and specialised genera, and together with certain closely ralated tribes they probably represent a separate line of evolution. The work of Schilder and Schilder demonstrates that when adequate records are available, a valuable picture of the food and feeding specialisations of a group of predatory insects can be built up, based solely on isolated field observations published in the literature. Although very useful, however, it would be wrong to believe that such a picture constitutes a complete analysis of tie situation. There are a number of different ways in which the information provided by field observations may be quite misleading. The reports are sometimes inaccurate or wrong; Schilder and Schilder point out several instances where they believe the observer was mistaken. They can easily show a bias towards prey which is taken in exposed situations where the act of predation is more likely to be observed; prey taken at night or in concealed situations will tend to go unrecorded. 45.

More important, however, are two basic limitations. Firstly, field-feeding records provide little or no information about the relative food value of different prey to the predator. Secondly, where there are indications of some degree of specificity in the relationship between predator and prey, they reveal nothing of the exact nature of this specificity and the way in which it is achieved. Forbes (1880) dissected adult specimens of three aphidophagous species and examined the gut contents, with surprising results. One specimen of Coleomegilla maculata Timb. had "its entire alimentary canal closely packed with pollen cells (dandelion)". The food of a specimen of Coccinella 9-notata Herbst. was estimated as 90% fungal spores and only 10% aphids, and that of a Hippodamia convergens Guer. as 75% fungal spores with a mite, an aphid and some Compositae pollen. It seems that these specimens were collected in conditions of aphid scarcity, for he suggests that "if they (predacious insects) can live on the soft parts of plants when animal food is scarce, their numbers will be maintained at a far higher figure than would be possible if they depended on 46.

animal food alone". From a later examination of the gut contents of 39 adults of six aphidophagous species (Hippodamini and Coccinellini), Forbes (1883) concludes that "the food of Coccinellidae seems to be remarkably simple and uniform, consisting wholly of spores of the lower cryptogams, pollen grains and plant-lice, varying but little from one genus to another". In the specimens he dissected, he estimated that aphids represented only about i of the food, almost of which was fungal spores. His general conclusion is clearly based on insufficient evidence, but nevertheless his results are interesting. With regard to specimens collected in a cornfield, the food of which he estimated to be 56% fungal spores, 13% pollen and only 14% aphids, he suggests that "aphids had been very much more numerous some time before, and it is possible that, as a consequence of this decrease of food, and the increase of the Coccinellidae themselves, the latter had reached an excessive number, for which the supply of plant-lice was insufficient, and that for this reason they had resorted to fungi". In spite of his results, therefore, Forbes still retained the belief 47.

that aphids were the preferred food, and other food sources were used only when aphids became unavailable. The utilisation of emergency foods by Coccinellidae was also suggested by Schilder and Schilder. Their review indicates that almost every tribe has a main food, and tney interpreted the minority foods recorded either as definite observational errors, or as emergency foods taken only in certain circumstances, such as when the main food is lacking or in artificial rearing conditions. A few examples will illustrate the types of food sources that adult Coceinellids may utilise if their normal food is scarce or unavailable, Schroder (1904) records many Coccinella 7-punctata L. feeding on the terminal shoots of young silver-fir, to such an extent as to cause permanent deformation of the plants. Hawkes (1920) found this species eating the pollen of stinging nettles in the summer, and the exudation from bean plants and Jerusalem artichokes in the autumn. He also observed Adalia bipunctata L. in the summer "apparently eating the juice of raspberries and split plums". Watson and Thompson (1933) observed Leis 48.

conformis Boisd. feeding at the extra-floral nectaries of Crotalaria striata, and also on pollen of saw-palmettoes terminal buds of scrub-oak, and sap from wounds caused by pruning of tangerine trees. Putman (1955) states that the mite-eater Stethorus punctillium Weise may feed on aphids and on peach foliar gland secretions when starving, but th,3 e focds do not allow oviposition. Aphidophagous species may utilise other insect food; Winkler (1952) reports Coccinella 7-Lunctata feeding on large numbers of su.badults and adults of the mayfly Ameletus iE2pinatus in the High Tatra mountains Adalia bipunctata may fcod on mites (Ross and Robinson 1922), on Psyllidae (Kanerlo 1940), and on Aleyrodidae and Coccidae (Hepper 1959). Coleomegrilla maculates DeG. is known predator of European corn borer eggs (Conrad 1959), and eggs of the cotton worm Prodenura litura P. in Egypt may be eaten extensively by Chilomenes vicina Muls (Bishara 1934) and Coccinella 11-punctata aegyptiaca Reiche (Ibrahim 1955) during periods of aphid scarcity. Ibrahim found that the latter species could not grow or reproduce on a diet of Prodenia eggs. Cannibalism is a frequent occurrence among 49.

Coccinellidae. In those which lay their eggs in batches, the first food of the young larva may frequently be unhatched or infertile eggs, and such feeding can greatly prolong the life of the larva and increase its serdting capacity when food is scarce (Banks 1956a). Cannibalism also occur:; among the larvae, and larvae as well as adults may eat prepupae and pupae when the normal prey is unobtainable (Hawkes 1920). The nourishment from cannibalism is apparently not a perfect one, at least in some cases (Tan 1934). An attempt to combine field observations with laboratory feeding tests was made by Kanervo (1940), working on the Coccinellidae of alder forests in Finland. He roughly classified the food taken by individual species into three categories; main food, subsidiary food and emergency food, basing his conclusions partly on field observations and partly on the degree of attraction shown to the food in the laboratory. His results indicate specificity at the tribal level of the kind shown by Schilder and Schilder, and also reveal some interesting food specialisations; for instance, Calvia 14-guttata 1. fed mainly on aphids whereas Calvia 15-guttata F. in 50.

the same habitat appeared to prefer Chrysomelid larvae. Hodek (1956, 1957, 1962) proposed that a careful distinction be made between 'acceptability' and 'suitability' of food in predacious insects. A wide range of food may be accepted by a predator under various conditions, especially at low prey densities, but only certain foods, which Hodek terms 'essential', are suitable for full development of the larva and maturation of the adult insect. Other foods which are accepted, designated as 'alternative' foods, may serve to carry the predator through a period when the essential food is scarce or unavailable, but in some cases they may be completely unsuitable or even harmful. The apparent polyphagy of many aphidophagous Coccinellids is partly due to this failure to make a distinction between acceptability and suitability, which is only possible if the effects of different foods are compared in the laboratory. Thus Hodek (1956, 1957, 1959) found that Aphis sambuci L. was lethal to both larvae and young adults of Coccinella 7-punctata L. Adalia bipunctata L., however, was able to develop on Aphis sambuci. Hodek also found that Semiadalia 11-notata Schneid. was adversely 51.

affected by feeding on Pergandeida medicaginis Koch., which was however a suitable prey for C. 7-punctata. According to Moursi and Kamal (1946), Leis conformis Boisch. will accept and feed on a wide range of prey, but it is incapable of completing its development on many of them. Okamoto (1961) found that Brevicoryne brassicile L. was more suitable as food for Coccinella 7-punctata than Rhopalosiphum prunifoliae. A.F.G. Dixon (1958) has recorded that Megoura viciae Buckt. is poisonous to Adalia 10-punctata L. Many species of Coccinellid have been reared in laboratory or insectary conditions and the larval develop- ment time and the fecundity and longevity of the adults recorded. In order to estimate the value of various species as predators, many workers have recorded the numbers of prey eaten by larvae and adults (see, for example, Clausen 1916, Palmer 1917, Smee 1922, Johnssen 1930, Cressman and Dumestre 1930, Putman 1957, Nielson and Currie 1960, Kaddou 1960, Koide 1961, Okuno 1961), but such data are of limited application. Information on the relative suitability of different prey species as food for Coccinellids is limited to the few examples 52.

given in the last paragraph, which however indicate that there can be considerable variation in the suitability of accepted prey. One subject which has received attention from several workers, again with economic aims chiefly in view, is the possibility of rearing Coccinellidae on artificial diets (Szumkowski 1951, Smirnoff 1958, Smith 1960, Chumakova 1962). Smith (1965 a, b and c), in particular, has succeeded in rearing certain species on various pollens, on synthetic diets based on yeast, cereals, beef and liver, and on dried, 179wdered aphids with water. Coleomegilla maculata lengi Timb. was particularly successful on artificial foods of all kinds, and even survived better and produced larger adults on dried powdered aphids than on living aphids. Smith (1965 a) suggests that this may be due to the need for some impurity, such as pollen, which may be present in the dried aphid diet. C. maculata could be reared to the adult on pollen alone, and an aphid-pollen mixture gave much greater survival and larger adults than aphids alone, so pollen may be an important part of the natural diet of this species, not just an emergency food. 53.

Coccinella spp. were considerably less successful on artificial diets, but Coccinella transversoguttata Kby. could be reared on dried powdered aphids. The use of dried, powdered aphids instead of living prey greatly facilitates measurement of the quantitative and qualitative effects of food on the predator. Smith (1965 a, b, c) has measured food intake, growth rate and growth and reproductive efficiencies for Coleomegilla maculata lengi Timb. and Anatis mali Auct. fed on dried aphids of different species. Such data, however, have no direct application to the effects of living food, for the use of dried powdered aphids eliminates many of the factors which influence and control predator-prey relationships under natural conditions. There are a few observations in the literature of aphid species which are less readily attacked than others. Hawkes (1920) claimed that he found three species of aphid, Macrosiphum aconitum Van der Goot, Hyalopterus arundinis P. and Aphis fabae L., unacceptable to Adalia bipunctata L. Smee (1922) succeeded in feeding A. fabae to Adalia bipunctata and Coccinella 7-punctata L. in the laboratory, but "not more than one or two were 54.

eaten at one time and these were only partially devoured and discarded", and he never observed feeding on this aphid in the field. Banks (1955) however, found that Adalia bipunctata was the commonest Coccinellid species associated with A. fabae on Vicia faba. Campbell (1926) observed that lalthough lady beetles do not attack pea aphids (Acyrthosiphon pisum Harris) as readily as other species of aphids, they can still be found in infested pea fields in large nui:lbers, feeding on pea-aphids". The first part of this statement seems to be unfounded. It can be said that in general aphidophagous Coccinellids accept as food any aphid species which they encounter. The varying suitability of different species as food for aphidophagous Coccinellidae is one aspect of the problem of specificity; the part played by habitat selection is another. We have seen that in certain other predators, Anthocoridae and Neuroptera for example,"selection of a particular oviposition site by the adult can play a considerable part in determining the degree of specificity. The factors involved in determining the oviposition site and habitat of aphid-eating Coccinellids are rather obscure. Eggs are not necessarily laid in strict relation 55.

to colones of prey. Banks (1956) investigated the correlation between numbers of Coccinellid egg batches and colonies of Aphis fabae on Vicia faba. During the early stages of an infestation when there were still many uninfested plants, any association between Coccinellid egg batches and aphid-infested stems was attributable to chance; many egg-batches were laid on uninfested stems, suggesting that the Coccinellids did not need the stimulus of the immediate presence of aphids in order to lay their eggs. When the infestation had reached an advanced state, however, the egg-batches tended to occur more frequently on the stems with most aphids. Banks considers that this was due to the concentration of adult Coccinellids on well-infested stems where they would tend to stay to feed on the aphids and lay their eggs. A.F.G. Dixon (1959) records the association of Adalia 10-punctata L. with Eucallipterus tiliae L. infesting about 20% of the leaves of a lime tree (Tilia x vulgaris Hayne). Of 37 egg-batches, comprising a total of 526 eggs, 81% were laid within 3.75 cm. of an aphid. Dixon suggests that Coccinellid eggs which 56.

are laid some distance from aphids, or on plants bearing few or no aphids, are probably laid soon after the Coccinellids have left a locality where aphids are present. He considers that the adults of Adalia 10- punctata oviposited close to aphids because they were less active through being well-fed, and consequently they remained in the vicinity of their own food supply. It therefore seems probable that oviposition behaviour in aphidophagous Coccinellids does not generally involve response to a stimulus from the aphids. The probability that a larva hatching from the egg will be, close to a supply of food may perhaps depend entirely on the fact that the adult also feeds on aphids, and for this reason is more likely to be close to aphids when it lays its eggs. Coccinellidae do, however, show certain habitat wefereneea, although these oTten- appear rather Indefinite. Hodek, Stary and Stys (1962) for example, record that in Czechoslovakia Coccinella 7-punctata and C. 5-punctata are typical of fields, Adalia bipunctata is rare in fields but common on trees and shrubs, and Propylea 14- punctata lives in both kinds of habitat. The position 57.

has been clarified somewhat by Iperti (1961, 1965), who has recorded the distribution of seven species of aphidophagous Coccinellidae in the South of Prance. The species studied (Iperti, 1965) were the usually monovoltine Coccinella 7-punctata, Semiadalia (= Adonia) 11-notata Schneid. and Adalia bipunctata, and the polyvoltine Adonia variegata Goeze, Propylea (= Harmonia) 14-punctata L., Harmonia conglobata L., and Coccinella 10-punctata L. In order to distinguish between the acceptability of an aphid species as food by the adult and its suitability for reproduction and larval development, Iperti recorded the numbers of larvae and pupae separately from the numbers of adults. Owing to the almost ephemeral nature of the aphid infestations he used ecological units which he termed 'occasional biotopes', characterised by altitude, time of year, species of aphid, species of aphid host plant, and duration of the aphid infestation. He sampled infestations of 19 different spcies of aphid throughout two seasons and analysed the collection data with respect to aphid species, mean height and type of vegetation and month of year. 58.

The adults showed an evident lack of food special- isation. Larvae and pupae were more restricted in their range of prey but only one species, Semiadalia 11-notata, was truly specific in the sense of feeding on only one species of aphid (Aphis fabae). The most significant results, however, were obtained when the height and type of vegetation were considered. The different Coccinellid species showed clear preferences for certain levels of vegetation; the larvae and pupae especially showed a tendency to be restricted in their distribution, indicating a degree of oviposition site selection by the adults. Thus Coccinella 7-punctata was found mainly on aphid infestations on low plants (0-50cm.); so also was Semiadalia 11-notata, the larvae of which were further restricted to infestations of A. fabae. Propvlea 14- punctata and Adonia variegata were more widespread in their distribution but nevertheless showed a marked preference for the shrub level of vegetation (50cm. - 2m6 ); but P. 14-punctata appeared to prefer cultivated plants whereas A. variegata was found more often on wild plants. Harmonia conglobata and Coccinella 10-punctata tended to 59.

be restricted to aphid infestations at shrub or tree level (above 50 cm.). Adalia bipunctata adults were relatively widely distributed but the larvae were almost confined to the tree level (above 2m.). Iperti also describes how seasonal fluctuations in the climatic conditions, affecting the successive appearance of different aphid infestations, influence and control the distribution of the Coccinellids and may mask their specific habitat preferences, especially in the summer months when aphids are less abundant. From his work a complex picture begins to emerge of the way in which the interacting factors of climate, aphid distribution and predator behaviour determine the distribution of. a Cocctnellid species and influence its degree of specificity. 60.

SECTION 2.

THE SUITABILITY OF DIFFERENT APHIDS AS

FOOD FOR

ADALIA BIPUNCTATA L. & COCCINELLA 7-PUNCTATA L. 61.

INTRODUCTION

The main purpose of the experimental work was to investigate differences in the suitability of various species of aphid to two common aphidophagous Coccinellids, Adalia bipunctata L. and Coccinella 7-punctata L. These two species are recorded by Schilder and Schilder (1928) as feeding on a wide range of aphid species, but recent work (Hodek 1956, 1957, 1959) suggests that in terms of suitable prey they may be more restricted. There also seems to be an interesting difference in habitat selection behaviour (Hodek 1959, Iperti 1965). Work which compares the food relations of these two species is therefore of some interest and may help towards a better understanding of the nature of specificity in this important group of predators. 62.

REARING TECHNIQUES AND GENERAL PROCEDURE.

Laboratory cultures of the two Coccinellid species used in the experimental work, Adalia bipunctata L. and Coccinella septempunctata L., were maintained in a constant temperature room at 20° with a 16 hour photo- period. All experimental work was done in these conditions. The standard rearing units were 3 in. x 5* in. x 2* in. high polystyrene boxes, ventilated by two 1 in. diameter holes cut in the lid and covered with bolting silk. (Fig. 1 a). A i in. high partition was cemented across the floor of each box to separate a compartment in which damp cotton wool was placed. A piece of corrugated cardboard, cut and shaped to fit the floor of the box and extend up one side, provided an absorbent substrate and also a surface suitable for oviposition. Food was provided daily in the form of cuttings from aphid- infested plants, inserted so that the cut ends were embedded in the cotton wool. In this way the plant was kept fresh for the aphids, which remained on it for about 24 hours. Acyrthosiphon pisum Harris was the 63.

aphid most commonly used in feeding the stock cultures, as it was comparatively easy to rear this species in the large quantities necessary for feeding Coccinellids. Up to six pairs of Coccinellids were kept in each rearing unit. These boxes were also used in experiments to determine fecundity, when a maximum of three pairs of Coccinellids were kept in each unit. Eggs were collected from the cultures by removing the corrugated cardboard and replacing it with a fresh piece. A very large proportion of the eggs was laid on the corrugated cardboard, in spite of the presence of fresh leaves bearing aphids, which formed a natural site for oviposition. Banks (1954) records Coccinellid eggs being laid on a cigarette carton in the field. Newly-hatched larvae were left for about 24 hours until they were about to disperse from the egg-cluster, and were then separated to avoid cannibalism. The apparatus most commonly used for rearing the larvae is shown in Fig. lb. The unit comprised 12 individual rearing cells in two rows of six. Each cell consisted of a 1 in. diameter by 2 in. high Perspex cylinder, resting FIGURE 1.

(a) Diagram of polystyrene box adapted for breeding

Coccinellids.

(b) Individual rearing cells for Coccinellid larvae. Corrugated Young bean cardboard plant

Cotton wool

(a)

Rearing cells

641/%0 . rot Nor

Reser voir F ilter paper wick

( b ) 66.

on a leaf laid with the undersurface uppermost on wet filter paper. The filter paper formed a wick leading up from a reservoir of water beneath the cells. In this way leaves of the appropriate aphids' food plant were kept turgid for 3-5 days. The lid of each rearing cell consisted of bolting silk stretched across a lf in. diameter galvanised iron ring, the weight of which was sufficient to keep the lower rim of the perspex cylinder closely applied to the leaf. It was generally necessary to renew the leaves once or twice during the development of a batch of larvae. The apparatus just described was also used in experiments on larval feeding. Adult beetles were collected in the field in the warmer months of the year and added to the laboratory cultures to revitalise them. In the course of the work the following species of aphid were cultured in the laboratory:- Acyrthosiphon pisum Harris (green variety) Acyrthosiphon pisum Harris (pink variety) Aulacorthum circumflexum Buckt. 67.

Myzus persicae Sulz. Megoura viciae Buckt. Aphis fabae Scop. The above species were all cultured on broad beans, Viola faba (dwarf Sutton) in good quality soil in polythene pots, four plants to a pot. Very good cultures of Myzus persicae were reared on Viola faba kept at a constant temperature of 200, although this aphid is relatively uncommon on leguminous plants in the field. In addition the following aphids were cultured on Brussels sprouts, Brassica aleracea gemmifera (Irish Elegance). Myzus persicae Sulz. Brevicoryne brassicae L. The other two species of aphid used, Aphis sambuci L. and Microlophium evansi Theo. were collected from field infestations. A field population of Aphis fabae on Euonymus europaeus was also used. The rearing techniques developed for particular experiments are described appropriately in the text. THE EFFECT OF DIFPMEET APHID FOODS ON TIE DEVELOPM2NT OF THE LARVAL PREDATOR

(A) Adalia bipunctata

In comparing the effects of different aphid foods on the development of larvae of A. bipunctata the following criteria were used: (1) the rate of development, (2) the weight of food eaten, (3) the mortality during development, and (4) the weight of the adult at emergence.

(1) The Rate of Development. Newly hatched larvae about to disperse from the egg clusters were divided into batches of 12, and placed individually in rearing cells. Each batch was given a different species of aphid, and the increase in live weight of each larva was measured daily. A CAHN-GRAM electro micro-balance was used to weigh larvae of 1st. and 2nd. instars. 3rd. and 4th. instar larvae were weighed on a beam balance accurate to 0.1 mg. The following aphid species were used: 4his fabae 69.

Aulacorthum circumflexum Myzus persicae Acyrthosiphon.zisum Microlophium evansi Fig. 2 compares the development of larvae fed on these species. It can be seen that larvae of A. bipunctata developed equally quickly on Aulacorthum circumflexum, Myzus persicae, ApirthosiRlhon pisum and Microlophium evansi, but the other two species, Aphis fabae and Aphis sambuci 6Ignifigant1y sliJcved the-- rate of development. Data for development time, measured from the start of feeding, are given in Table 1. FRESH WEIGH T OF LARVA(P AGNS.) FIGURE 2 Development ofAdaliabipunctatalarvaeondifferent aphidfoods. AGE OFLARVA(HOURS) 71.

SPECIES OF DEVELOPMENT TIME (DAYS) APHID USED LARVAL INSTARS TOTAL PREPUPA PUPA AS FOOD 1 11 111 1V Aulacorthum + circumflexum 2.1 2.0 2.4 3.3 9.6 - 0.6, 1.7 5.9 1 My zus persicae 2.1 2.0 2.3 3.7 10.3 ± 0. 8 1.8 5.6 Aphis fabae (Vicia) 3.0 2.8 3.0 5.4 14.3 ± 1.9 1.6 5.8 Aphis fabae (Euonymus) 3.3 2.9 3.0 4.8 13.9 1.2 1.7 5.8 Acyrthosiphon pisum (green) 2.2 1.9 2.5 4.1 10.7 -0.9 1.5 5.6 Acyrthosiphon pisum (pink) 2.4 1.7 2.3 4.2 10.7 1.1 1.7 5.6 Aphis sambuci 2.7 2.4 3.3 5.3 13.7 1.5 1.7 5.7 Microlophium evansi 2.2 1.8 2.7 4.0 10.7 0.8 1.3 5.9

TABLE 1 - Development time of A. bipunctata larvae on different aphid diets.

This table also includes development times of larvae fed on A. fabae collected from a field infestation of Euonymus. The different host plant did not alter the food value of A. fabae which was equally inferior whether bred on Euonymus europaeus or on Vicia faba. 72,

Table 1. shows that larvae of A. bipunctata developed at the same rate on green and pink varieties of Acyrthosiphon pisum. It also shows that durations of the prepupal and pupal stages were not altered by the species of aphid fed to the larvae.

(2) The Weight of Food Eaten. The live weight of aphids taken dbily by developing larvae was estimated as follows. Young first instar larvae were placed in separate rearing cells. Two sets of small, polythene-stoppered vials, numbered to correspond with the rearing cells, were used in weighing the aphids. Weights of the vials were recorded at the start, and checked at the end of the experiment. Aphids of an appropriate size (small for small larvae) and quantity to supply the larvae with food for 24 hours were placed in one set of vials. The weights of vials + contents were recorded, and the aphids were presented to the larvae in the rearing cells. After 24 hours the aphids and aphid remains left in each cell were returned to the correspond- ing vial, which was then reweighed. A weighed quantity of fresh aphids was then added to each cell from the 73.

second set of vials. Thus, the two sets of vials were used on alternate days to weigh the food presented to the larvae and the material remaining at the end of a 24 hour period. In this way a measure of the weight of food taken daily was obtained by subtracting one day's readings from those of the preceding day. A correction was applied to the data to compensate for gain in weight of the living aphids during the 24 hour period. Batches of control aphids were treated in the same way as those fed to the predators, weighing before and after the 24 hour period, and from them an 'aphid growth factor' was obtained: weight of control aphids after 24 hour period 'Aphid growth factor' = 1 x weight of control aphids before 24 hour period This involves the necessary assumption, which may not be valid, that the predators feed at a uniform rate through- out the 24 hour period. The figures obtained in the experiments were multiplied by the appropriate 'aphid growth factor' for the species of aphid concerned, to obtain a better indiction of the weight of food actually 74 .

taken by the predator. In spite of this procedure the data obtained for weight of prey eaten can only be regarded as an estimate, as no allowance can be made for water loss from the aphid remains. However, the results of these experiments provide a valid comparison of the relative weights of different species of aphids taken by A. bipunctata larvae during development. The following aphid species were used as prey:

Aphis fabae Aulacorthum circumflexum Myzus persicae Acyrthosiphon pisum Megoura viciae Brevicoryne brassicae

Fig. 3 compares the weights of food eaten when A. bipunctata larvae were given four of these species. It can be seen that, correlated with the slower rate of development of larvae feeding on A. fabae, there was a slower rate of uptake of food. The fact that A. fabae was less suitable as food for A. bipunctata larvae is Mean Wet Weight of Aphids Eoten (Mgms.) FIGURE 3 Rate ofUptakeDifferentAphidsbyLarvaeAdalio bipunctotoL. Age ofLarva(Hours) 76.

therefore at least partly attributable to the inability of the larvae to feed on it as quickly as on other aphid species. Further consideration will be given to this point later. In spite of this difference in rate of food intake, however, the total weight of food eaten during larval development was similar for all aphids. All larvae fed on Iviegoura viciae died before they were two days old, in the first instar. Later experiments show that this aphid is eaten, but is toxic. Six larvae were given Brevicoryne brassicae as food, and of these only two became adult. Of the rest, one died in each instar. B. brassicae could not be weighed accurately because of its waxy secretion, and no valid data for weight of food eaten could be obtained. The two larvae which completed their development on B. brassicae took 21 and 23 days respectively from hatching to pupation, compared with 9-14 days for development on the other aphid species (Table 1). The fresh weights of food taken by each larval instar fed on the other four species of aphids are shown in Table 2.

77.

APHID SPECIES FRESH WEIGHT OF FOOD TAKEN (MG.) USED AS PREY 1 11 111 1V TOTAL

Aulacorthum circumf1exum 0.13 2.30 11.45 48.75 62.15 ± 6.60 My zus persicae (.32 2.32 10.88 56.85 69.14 ± 11.20 Acyrthosiphon pisum 3.22 10.73 49.80 63.75 ± 4.64 Aphis 57.81 ± 6.32 fabae 0.35 1.97 9.73 46.25

MEAN 0.20 2.45 10.70 50.41 63.76

i of food taken in each instar 0,31% 3.84% 16,78% 79.07%

TABLE 2 - Weight of food eaten by A. bipunctata larvae.

It is evident that the weights of food eaten by larvae fed on these four species of aphids do not differ significantly. It is of interest that about 80% of the total food eaten during larval development is consumed by the fourth instar. (3) Mortality during larval Development. Mortality is frequently used as an indication of

78.

the suitability of a food. In this work, however, it was difficult to get reliable figures for mortality. There were considerable differences between the mortalities of larvae from different egg batches given the same food. Mortalities of larvae fed on six aphid species are given in Table 4, and confirm the relative unsuitability of A. fabae and A. sambuci. The mortalities obtained with A. fabae and M. persicae are based on the development of large numbers of larvae and deserve further consideration. Table 3 shows the way in which mortality was distributed through the instars.

SPECIES OF NUMBER NUMBER DYING IN EACH INSTAR TOTAL MORTALITY APHID USED OF I 11 111 1V Pupa AS FOOD LARVAE

Aphis 2 fabae 174 16 9 8 13 48 27.6% Myzus 17.8% persicae 107 11 3 1 1 3 IO

TABLE 3 - Mortality of A. bipunctata larvae fed on A. fabae and M. persicae.

In both cases, the number of deaths was highest in the first instar, but with A. fabae, mortality was also 79.

high later in larval development. In contrast, larvae fed on M. persicae had a high probability of reaching the adult stage once they had attained the second instar.

(4) Weight of the Adult at Emergence. The two aphid foods which decreased the rate of development, A. fabae and A. sambuci, also resulted in smaller adults. (Table 4),

SPECIES OF MORTALITY DURING MEAN WEIGHT OF APHID USED LARVAL DEVELOP- ADULT AT AS FOOD MENT (%) EMERGENCE (MG.)

My zus persicae 17.8 11.8 ± 1.6 Aulacorthum circumflexum 16.7 11.9 ±1.9 Aphis fabae 27.6 8.4 ± 1.6 Aphis sambuci 25.0 8.0 ± 1.4 Acyrthosiphon pisum 18.8 12.4 t 2.1 Microlophium evansi 9.1 12.4 ± 2.3

TABLE 4 - The effects of different aphid diets on the mortality during larval development and on the size of adult in A. bipunctata. 80.

Two adults which were reared from larvae fed on Brevicoryne brassicae were very small, weighing 5.1 mg. and 6.1 mg. respectively. Both died within a week after emergence. Table 5 shows the initial weights of each instar and the effective increase in weight during each instar for larvae fed on different aphids.

SPECIES OF LARVAL INSTARS MEAN WT. OF AIHID USED 1 11 111 1V ADIZT AT AS FOOD iw inc iw inc iw inc iw inc EMERGENCE (MG) Myzus persicae 0.1 0.3 0.4 0.8 1.2 2.6 3.8 8.0 11.8 Aulacorthum circumflexum 0.1 0.3 0.4 1.1 1.5 2.5 4.0 7.9 11.9 Aphis fabae 0.110.2 0.3 1.1 1.4 2.4 3.8 4.6 8.4 Aphis sambuci 0.1 0.3 0.4 0.9 i.5 2.5 3.8 4.2 8.0 Acyrthosiphon pisum 0.1 0.4 0.5 0.9 1.4 2.6 4.2 8.4 12.6 Microlophium evansi 0.1 0.3 0.4 0.7 1.1 2.8 3.9 8.5 12.4

TABLE 5 - Comparison of the initial weight of each instar Liwarl....211122e112211Ye1-a212E2211ELral each instar (inc) for larvae of A. biulictata fed on different aphids. 81.

The initial weight of the 4th instar is similar irrespective of the aphid used as food. The lower weights of the adults obtained from larvae fed on A. fabae and A. sambuci are due to the failure of the 4th instar to reach full size. 82.

(B) Coccinella septempunctata.

The increase in weight of C. 7 - punctata given different aphid foods is illustrated in Figs. 4 and 5. It can be seen that for this species, the rate of development of larvae fed with A. fabae was as fast as with any other aphid (Fig. 4). Larvae of C. 7 - punctata were also able to develop on Megoura viciae, the aphid which was toxic to A. bipunctata, but at a slower rate than on the other aphids. Larvae fed on Brevicoryne brassicae developed very slowly (Fig. 5), and this did not appear to be associated with Brassica as the aphids' food plant, since the rate of development of larvae fed on Myzus persicae was the same irrespective of whether the aphids were reared on Brassica or on Vicia. Data for development times are given in Table 6. Also included in this table are data for a group of larvae fed on Aphis sambuci. 50

M. persicoe 40 A. pisum

A. foboe

) ms Mg

( 30 rva M. viciae f La o ht ig We 20

10

100 200 300

Age of Larva ( Hours)

FIGURE 4.

Development of Coccinella 7-punctata on different aphid foods- 1 50

40

) ms. Mg

( 30 rva f La o ht ig We

h 20 Fres

10

100 200 300

Age of Larva ( Hours )

FIGURE f Development of Coccinella 7-punctata on different aphid Foods - 2 85.

SPECIES OF DEVELOPMENT TIME (DAYS) APHID USED LARVAL INSTARS TOTAL PREPUPA PUPA AS FOOD I II III IV Myzus persicae 3.9 2.0 2.3 1 4.8 13.0±0.7 2.0 6,0 Aphis fabae 3.9 2.0 2.9 5.0 13.7±0.6 1.8 6.0 Acyrthosiphon pisum 3.9 2.0 2.5 4.8 13.2±0.8 1.5 6.4 Megoura viciae 4.4 2.4 3.1 5.1 14.8±0.7 1.7 6.3 Brevicoryne brassicae 5.0 2.9 3.1 5.1 16.1±0.6 1.3 6.0 Aphis sambuci 4.3 3.1 4.3 8.5 19.5±3.7 1.3 6.0

TABLE 6 - Development time of C.7-punctata larvae on different aphid diets.

The rate of development of larvae feeding on Megoura viciae and Brevicoryne brassicae was only significantly decreased in the let and 2nd instars. Conversely, larvae feeding on Aphis sambuci had incr,:asing difficulty in developing on this aphid as they grew older. The adverse effect of A. sambuci on C. 7-punctata is recorded by Hodek (1956, 1957). Data on mortality during larval development, and mean weights of adults at emergence, are summarised in

86.

Table 7.

SPECIES OF MORTALITY DURING MEAN WEIGHT OF APHID USED LARVAL DEVELOP ADULT AT AS FOOD -TENT (%) EMERGENCE (MG.)

My zus 5.9 36.4 ± 5.0 -yr persicae I I Aphis ± 2.5. fabae 5.9 36.3 Acyrthosiphon 37.2 ± 4.4 11 pisum 25.8 Megoura viciae 14.7 33.5 ±4.2 Brevicoryne brassicae 26.1 30.9 ± 2.9 Aphis 2.2 sambuci 50.0 18.4 ±

TABLE 7 - The effect of different aphid diets on the mortality during larval development and on the size of adult in C. 7-punctata. Values for adult weight which are not significantly different (P,01.05) are bracketed together.

Apart from the unexpectedly large mortality of larvae fed on Acyrthosiphon pisum, the figures for mortality and adult weight confirm the differences in suitability as indicated by the rates of larval development (Table 6). 87.

THE EFFECT OF DIFFERENT APHID FOODS ON THE FECUNDITY OF THE ADULT PREDATOR

(A) Adalia bipunctata. For records of fecundity a large number of larvae were reared simultaneously. The adults were weighed and sexed on emergence, and separated, three pairs to each box. They were provided daily with an excess of food, and eggs were collected and counted on alternate days. Fig. 6a compares the fecundity of adults fed on Aphis fabae and onNyzus persicae, the larvae having been reared on the same food as the adults. 12 pairs of beetles were fed on each species of aphid. Feeding on A. fabae reduced the fecundity by more than half. In addition the eggs laid by females feeding on A. fabae were significantly smaller (both in length and breadth), and their fertility was considerably reduced. (Table 8). Comparison oftheFecundityAdaliabipunctatafedonAphis fabaeandMyzuspersicoe. FIGURE 6 MEA N NUMBEROF EGGSPE R FEM

A ALE B

23 I t larvae rearedonoppositefood. larvae rearedonsamefoodasadult. DAYS AFTEREMERGENCE DAYS AFTEREMERGENCE B Myzus persicae Aphis fabae

89.

SPECIES OP APHID

USED AS FOOD MYZUS PERSICAE APHIS PABAE

Mean emergence wt. of females (mg) 10.7 8.9 Mean longevity of females (days) 76.2 83.0 Mean total eggs laid 676.2 249.6

Mean eggs per day 6.9 2.6

Maximum eggs per 18.0 7.7 day Mean egg size - length 1.04mm 0.99mm Mean egg size - breadth 0.48mm 0.46mm

% fertility 89.4 55.9

TABLE 8 - Fecundity data for A. bipunctata fed on M. persicae and A. fabae.

In order to investigate the effect of the food of the larva on the fecundity of the adult, further batches of larvae were reared on these two aphid species, the emerging adults being fed on the opposite food to that given to the larvae. Thus adults developing from larvae reared on A. fabae were given M. persicae, and adults from M. persicae-fed larvae were given A. fabae. Fig. 6b compares the fecundities of these two groups of 90.

adults, and the results of the experiment are summarised in Table 9.

Larval food Aphis fabae Myzus persicae Adult food Myzus persicae Aphis fabae

Mean emergence wt. of females (mg) 9.5 10.7 Mean longevity of 92.2 78.8 females (days) Mean total eggs 1068.2 360.0 laid Mean eggs per day 9.8 4.0 Maximum eggs per 20.1 7.8 day % fertility 95.5 69.2

TABLE 9 - Fecundity data for -1. bipunctata fed on M. persicae and A. fabae, larvae having been reared on opposite food to adult.

Surprisingly, the food given to the larva did not affect the fecundity of the adult. In this experiment, however, the mean emergence weights of the adults obtained from larvae reared on the two species of aphid were closer than is usual, since females from larvae fed on M. persicae were lighter than in other experiments and females from larvae fed on A. fabae were heavier. The 91,

difference is, however, still significant at the 5% probability level. Acyrthosiphon pisum seemed to be a suitable food for adult A. bipunctata and eggs were laid, but fecundities were not recorded. Megoura viciae was toxic to adult A. bipunctata, all beetles fed on this species dying within seven days without ovipositing. 92.

(B) Coccinella septeinpunctata.

This species is liable to eat its own eggs even when aphid food is plentiful, so to record fecundity eggs had to be collected very frequently. Females were sometimes observed to lay a batch of eggs00 and then turn around and eat them immediately. Such behaviour made it virtually impossible to determine fecundities accurately. The number of eggs collected and recorded is believed, however, to be a large proportion of the number of eggs laid, and serves to give a comparison of the fecundity of C. 7-punctata fed on different aphid foods. In this experiment, all larvae were reared on Acyrthosiphon pisum. Emerging adults were weighed, sexed, and divided into four groups of twelve pairs each, so that the mean weights of males and females in each group were about equal. They were then placed in boxes, three pairs to each box. The four groups of adults were then given different diets, (1) Myzus persicae, (2) Aphis fabae, (3) Acyrth- osiphon pisumt and (4) Acyrthosiphon pisum + pollen. In (4),catkins of hazel, and later, willow, were suspended 93.

in the boxes, and replaced daily. The results of this experiment are illustrated in Fig. 7, and summarised in Table 10.

SPECI-uS OF Mean Mean Mean Maximum 04/0 longevity total eggs eggs fertility APHID USED of females eggs per per day AS FOOD (days) laid day

Myzus 84.97 persicae 86.4 315.3 4.6 9.5

Aphis 76.2 13.2 69.68 fabae 351.3 6.1

Acyrthosiphon 21.6 80.77 pisum 79.3 403.7 6.4

A. pisum + 25.1 86.31 pollen 84.2 591.3 9.0

TABLE 10 - Fecundity data for C. 7-punctata.

The only significant difference in fecundity at the 5% probability level was between adults fed on Myzus persicae and those fed on Lurthosiphon pisum + pollen. There was no well-defined peak in the oviposition of females fed on Myzus persicae, but laying was maintained over a longer period (Fig. 7). There was no significant difference in the size of the eggs laid by female C. 7-punctata fed on Aphis fabae and on Myzus persicae, but the fertility of eggs from females fed on A. fabae

24 24

Myzus persicae A phis fobae

12 16

20 40 00 80 20 40 00 82 Days after emergence Days after emergence

24 Acyrthos,phan pisum 24 Acyrthosq:thon pisum . pollen

a 16

0 41

8 8

20 40 60 80 2. 4r 60 Days after emergence Days after emergence

FIGURE 7 THE EFFECT OF DIFFERENT DIETS ON THE FECUNDITY OF COCCINELLA 7-PUNCTATA L. 95.

was significantly less than in the other three groups (Table 10). Beetles provided with pollen fed on it frequently in spite of the abundance of aphids. The peak laying period was extended in these beetles compared with that of beetles fed on Acyrthosiphonjoisum alone. More eggs were laid but the fecundity was not significantly different at the 5% level. Six weeks after emergence, at the peak of the oviposition period, all experimental adults were re- weighed. A comparison of the weights at emergence and the weights of male and female beetles at this time is given in Table 11. •

MiP2MIMENTAL Mean weights of Mean weights of MALES (mg) FEMALES (mg) GROUP At After At After emergence 6 weeks Diff. emergence 6 weeks (1)Myzus persicae 34;8 43.8 9.0 37.1 46.8 9;7 (2)Aphis 33.2 39;9 6;7 38:5 57.0 13:5 fabae • (3)Acyrthooiphon pisum 32:5 41.4 8;9 37.4 53.0 15;6

(4)A. pisum + 21;2 pollen 32:8 43;3 10.5 35;9 57.1

TABLE 11 - Comparison of the weights of male and female C. 7-punctata at emergence and six weeks after emergence. 96;

All beetles had increased in weight since emergence. The mean weight of females fed on Myzus persicae was now, however, considerably less than that of females in the other three groups. Noticeably less food was being taken at this time by females given Myzus persicae. The very small adults reared from larvae fed on Aphis sambuci were all dead less than two weeks after emergence (feeding on A. sambuci being continued). Adults reared from larvae fed on Brevicoryne brassicae were also given B. brassicae in the adult stage. This species seemed to be a suitable food for adult C. 7- punctata, and eggs were laid, but no records of fecundity were made. A small group of adults (4 males, 4 females), were given Megoura viciae in the adult stage, the larvae having been reared on A. fabae. These adults all died within eight days; but a group of adults (10 males, 10 females) fed on M. viciae, which had also been given M. viciae in the larval stage, were more successful. Ten were still alive after 30 days and two lived for 90 days, but no eggs were laid. 97.

APHIS FABAE L. AS FOOD FOR ADALIA BIPUNCTATA L.

It was of particular interest to find that Aphis fabae, a common prey of A. bipunctata in Britain, is relatively unsuitable as food for larvae and adults of this species. The nature of this unsuitability was further investigated. A quantitative comparison was made of Aphis fabae and Myzus persicae as food for the 4th instar larva of A. bipunctata. Groups of predator larvae were reared to 4th instar on each species of aphid. Open 3 in. x 1 in. tubes were used in the experiment. These were fitted with an internal scaffolding cut out of Bristol Board, to give a large absorbent surface area, and with a thin coat of 'Fluon' around the inside of the rim to prevent the insects escaping. The tubes were kept in desiccators at 50% relative humidity, and weighed at equilibrium. All weighings were done on a beam balance weighing to Os' mg. One-day-old 4th instar larvae, kept without food overnight, were weighed and placed in the tubes. Large numbers of each species of aphid were collected from 98.

the cultures, and weighed amounts of food, about 50 mg. per predator, were presented to the larvae in the tubes. The tubes were then replaced in the 50% R.H. desiccators and kept for 24 hours in a constant temperature room at 20°. Tubes with aphids alone were used as controls to ascertain the loss in weight of live aphids kept unfed for 24 hours under the conditions of the experiment. After 24 hours all experimental and control tubes were weighed. The Coccinellid larvae were removed from the tubes and weighed. Live aphids which remained were removed from the experimental tubes and weighed. The remaining contents of the experimental tubes, consisting of aphid remains, dead aphids and aphid exuviae, were examined and the aphid remains separated from the rest and weighed. The predators and live aphids from both control and experimental tubes, were killed and dried to constant weight over phosphorus pentoxide at 1000, and the dry weights obtained. The aphid remains were also dried and weighed. This experiment provided several pieces of quantitative 99.

information. The weight of aphids killed by the predator was estimated by subtracting the weight of aphids left alive in the experimental tubes from the weight of aphids present at the start of the experiment, after correcting for weight lost during the 24 hour period. The weight of food ingested by the predator was then calculated by subtracting the weight of aphid remains from the weight of aphids killed. An estimate of the proportion of each aphid attacked which was ingested, here termed the 'feeding efficiency', was thus possible:

Feeding _Dry wt. of aphids killed-dry wt. of remains x 100 efficiency Dry wt. of aphids killed

The 'growth efficiency' of 4th instar A. bipunotata larvae feeding on each aphid species was estimated from the increase in weight of the predator and the weight of food ingested:

Growth Increase in dry wt. of predator efficiency Dry wt. of aphids killed-dry wt. of-remains x100

A comparison of the values for 'feeding efficiency' and 'growth efficiency' for larvae fed on the two aphid species is given in Table 12. 100.

SPECIES FEEDING GROWTH OF APHID EFFICIENCY EFFICI3NCY

Myzus persicae 40.1 ± 5.0 44.9 ± 6.7

Aphis fabae 23.6 t 4.5 28.1 ± 3.2

TABLE 12 — Comparison of v-lues for 'feeding efficiency' and 'growth efficiency' for A. bipunctata 4th. instar larvae fed on M. persicae and on A. fabae.

Thus A. bipunctata larvae feeding on A. fabae consumed relatively less of each aphid which they captured, and the material which was ingested was apparently less nutritious than that obtained from M. persicae. It is also of interest that the ratio of wet weight to dry weight of larvae fed on A. fabae is significantly grexter than that of larvae fed on M. persicae (4.66 ± 0.22 as against 3.88 ± 0.16); i.e. larvae fed on A. fabae have a higher water content. There is a noticeable difference in the behaviour of A. bipunctata 1,,rvae on A. fabae. The active pumping of digestive fluids into and out of the aphid body, which is a feature of the feeding behaviour of Coccinellid larvae 101;

(Hawkes 1920, Banks 1957; Kaddou 1960), is lacking when the prey is A. fabae. The normal behaviour of 4th instar A. bipunctata larvae feeding on nymphs and adults of M. persicae involves (1) extra-oral digestion, with injection of digestive enzymes into the body of the aphid alternating with the sucking out of body fluids and partially digested m.terial, and (2), partial or complete ingestion of the solid remains left after (1) 4th instar A. bi punctata larvae feeding on A. fabae were only observed to do this when the prey is a young nynph. With older nymphs and adult A. fabae, the active pumping in (1) is replaced by a relatively pass.lve sucking at the wound. The corpse of thu aphid is then left in a more or less intact state, apart from being somewhat deflated. The remains of adult A. fabae left after feeding by 4th instar A. bipunctata larvae were examined by cutting lOngitudinal sections of deflated corpses and comparing them with sections of whole aphids. Salivary gland, gut and fat-body had all been removed by the predator, but muscles were unaffected. Embryos had in some aphids been broken up with some of the contents 102.

gone, but the embryos were mostly unaffected. In the brain and sub-oesophageal ganglion, many neurones had gone although the neurilemma were unaffected. There thus appeared to have been a limited degree of extra- oral digestion. It was also noticed that the cuticle of adult A. fabae is 14 to 2 times as thick as that of adult M. persicae. A thick cuticle would make the active pumping in and out of material by the predator more difficult, and also might prevent the final mastication and ingestion of the solid remains. There are two possible reasons why A. fabae might be less nutritious than M. persicae as food for A. bipunctata. Either the whole aphid is nutritionally deficient, or some dietary requirement is being left behind in significant quantities in the remains left by the predator. Proceeding on the latter hypothesis, it was thought possible that the dietary requirement might be an amino- acid. The amino-acid composition of A. fabae remains left after predation by 4th instar A. bilDunctata larvae was therefore determined by two-way paper chromatography, and compared with that of whole aphids, other species of aphid, and the predator itself. 103.

Samples of adult Aphis fabae, Myzus persicae and Megoura viciae, 4th instar larvae of A. bipunctata, and A. fabae remains after predation, were dried and crushed. 4 mg. of each sample were weighted out, and hydrolysed in sealed tubes with 6 N hydrochloric acid at 45° overnight. The water and acid were then evaporated off under pressure. The residue was washed, and taken up in 0.2 ml. of water with 10% iso-proparkol as a preservative. Spots were applied to paper using a 1 micro-litre micropipette. Four chi.omatograms of each s;:zple were run at 5. 10, 15 and 20 pi concentrations. Four chromatograms of a 0.01 i glycine solution at 2, 4, 6 and 8)111 concentrations were also prepared. Runs were made overnight, the first solvent being butanol 120/ acetic acid 30/distilled water 50, and the second solvent phenol 200/ammonia 1. After the second run the papers were dipped in ninhydrin. The overall densities of all spots were measured with an 3EL scantling densitometer. The amino-acid concentrations of the samples were calculated by comparison with the standard glycine solution. Mean values were obtained from the four chromatograms run from each sample.

104.

Table 12 shows the percentap-e composition of the samples. Certain amino-acids rith spots close together in the chromatograms have been grouped together.

A. bipunctata 4th instar EYZUS MEGOURA APHIS A. FABAE larvae PER3ICAE VICIAE FABAE RETIAINS Serine/Glycine 19.88 19.41 19.12 19.91 19.75 Arginine/Lysine /Histidine 15.33 16.98 16.71 14.74 15.63 Leucine/ Phenylalanine 14.30 14.37 12.63 15.24 13.83 Glutamic acid 11.29 10.61 14.23 13.21 12.91 Alanine 10.84 13.64 11.57 10.63 13.05 Aspartic acid 8.20 6.50 9.14 10.36 9.65 Valine/ Methionine 8.91 8.30 7.39 7.51 8.55 Threonine 5.22 6.25 4.75 4.40 3.73 Tyrosine 4.94 2.89 3.44 2.58 1.27 Cysteine 1.09 1.02 1.01 1.38 1.49

TABLE 13 - Amino-acid analysis: % Composition of Samples. 105.

It is apparent that both predator and prey are closely similar in their amino-acid composition, and also that no amino-acid is being left behind in significant quantities in the A. fabae remains. Taking 24% as the proportion of an adult of A. fabae consumed by A. bipunctata, the percentage of available amino acid consumed by the predator was calculated. The values obtained are given in Table 13. A relatively small proportion of the alamine present in the whole aphids is obtained by the pred-tor, but it is impossible to say whether this results in deficiency. 106.

Weight -eight Weight Percentage available- left consumed of available ( g) ( g) consumed.

Serine/Glycine 10.5 7.1 3.4 32.4 Arginine/Lysine /Histidine 7.8 5.7 2.1 27.1 Leucine/ Phenylalanine 8.0 5.0 3.0 37.6 Glutamic acid 6.9 4.7 2.2 31.9 Alanine 5.6 4.7 0.9 16.1 Aspartic acid 5.5 3.5 2.0 36.8 Valine/ Methionine 4.0 3.1 0.9 21.8 Threonine 2,3 1.4 0.9 39.2 Tyrosine 1.4 0.5 0.9 66.2 Cysteine 0.7 0.5 0.2 27.8

TABLE 14 - Relative amounts of amino-acids obtained from A. fabae by 4th instar larvae of A. bipunctata. 107.

A CMPARATIVESTUDY OF "HE FEEDING BErAVIOUR OF LARVAE OF ADALIA BIPUNCTATA L. AND COCCINELLA SEPTEM7UNCTATA L. ON FOUR APHID SPECIES.

It was decided that better understanding of the effects of different aphid foods on larval development might result from a study of the feeding behaviour of the larvae. Accordingly observations were made of 1st. and 4th instar larvae of the two Coccinellid species feeding on four aphid species; Myzus persicae, Aphis fabae, Acyrthosiphon pisum and Megoura viciae. The procedure adopted was to present aphids of known weight to the predator and record the time spent in feeding on the prey and the percentage weight of the prey eaten. Unfed let. instar ldrvae were used, approximately 12 hours after dispersion from the egg cluster. The 4th instar larvae were reared on the species of aphid used in the experiment, allowed 3-4 hours feeding after the 3rd./4th. instar ecdysis, and then left without food overnight. Each predator was used only once. The aphids used as prey wore selected by weight. 108.

Young nymphs weighing approx. 0.3 mg. were given to the 1st. instar predators, and nymphs of approx. 0.5 mg. to 4th. instar predators, the instar depending on the species of aphid. Each predator was presented with 20 aphids of almost identical weight, in an arena consisting of the lid of a 2 in. plastic pill-box, the sides of which were coated with 'Fluon'. The method used to obtain 20 prey of almost identical weight was to weigh one aphid of suitable size, then select nineteen more aphids of similar size by eye, and weigh the complete batch of twenty. According to whether the whole batch weighed more or less than twenty times the original aphid, aphids which appeared to be larger or smaller than average in the batch were replaced until the weight was correct to within 0.01 mg. Further batches of 20 aphids were weighed in the same way, of which two were dried to obtain dry weights the others used in the experiments. This method of selection was found to give aphids agreeing in weight to within 5%. All weighings were done on a CAHN-GRAM electo-microbalance on the 100 mg., 10 mg. and 1 mg. ranges (ie. weighing to 0.01 mg., 0.001 mg., and 0.0001 mg. respectively). 109.:

Each experiment was replic,ited about 10 times. Records were made of (1) the time taken to make an effective capture, (2) the number of potential prey encountered (ie., touched by forelimbs or mouthparts) before and including the first effective capture, (3) the time spent feeding, from insertion of the mouthparts into the body of the aphid until either complete ingestion of the prey or abandonment of the remains, (4) the wet weight of the aphid remains left, and (5) the dry weight of the remains. In addition observations were made of the method of capture of the prey and the behaviour of the predator whilst feeding. The time taken to ma'ce s.n effective capture, and therkmber of potential prey encountered in this time, give an indication of the efficiency of the predator in capturing its prey, but results obtained under highly artificial conditions such as these must be treated circumspectly; the behaviour of both the predator and its prey under natural conditions on the food plant may be very different. The time spent feeding and the weight of aphid remains left are less likely to be affected by the conditions of the experiment, and in 110; confirmation of this some records made of larvae feeding on aphids on their food plant agreed closely with the results obtained here. Once a capture had been made the rest of the aphids in the arena were removed to avoid disturbance of the predator whilst feeding. The remains of the aphid left after feeding were weighed immediately to get a value for wet weight before significant water loss occurred. From the wet weight and dry weight of the remains left after predation the percentage of wet weight and dry weight eaten were calc- ulated, using values for the wet weight and dry weight of the original aphid estimated as described above. The feeding behaviour of the 1st. and itth. inotars of each species of Coccinellid will be considered separately, following which certain comparisons will be made. Adalia bipunctata 1st. instar (Table 15) First instar larvae of Adalia bipunctata were relatively inefficient predators. They rarely made a capture at the first encounter and generally took several minutes to capture an aphid. The larvae found Acyrthosiphon pisum most difficult to capture due to the rapid escape reactions of this species. In exactly half of the observed captures of A. pisum, the predator caught, hold of a tarsus, and

was dragged around by its prey for some time before it managed to grip the body of the -phid.

No. of Mean time Hean no. Bean time % uet dry PREY before 1st of prey spent weight weight reps. capture encountered feedinc eaten eaten (mins) (reins) Hyzus persicae 10 3.6 2.5 128.0 05.8 63.1 Aphis fabae 13 2.6 1.5 379.0 30.4 12.7 Acyrthosiphon pisurn 12 13.9 4.5 116.3 80.6 79.0 Hegoura viciae 4 7.5 3.5 13.5 %NO

TABLD 15 - Feedinfl behaviour of 1st. instar larvae of Adalia bipunctata.

Megoura viciae w-s also a difficult aphid to catch, but once a capture had been liade the predator showed no reluctance TO begin feeding. Yet the prey was soon released (Table 15), and on examination the predator was found to be either dead or dying. M. viciae thus has a very rapid toxic effect on lst. instar larvae of A. bipunctata. 1st. instar larvae of A. bipunctata spent about 2 hours feeding on individuals of M. persicae and 112.

A. pisum, during which time they consumed the greater proportion of the prey. When the prey was A. fabae, however, feeding was continued for three times as long, and only a small proportion of the prey was taken in in this time.

Adalia bipunctata - 4th. instar (Table 16) 4th. instar larvae encountered as many aphids before making a capture as did lst. instar larvae, but the time before first crTture was generally less due to the greater speed of movement of the larger larva.

Mean time Mean no. Mean time 5 wet Vo dry No. of before 1st of prey spent weight weight reps. capture encountered feeding eaten eaten (mins) (mine Iiyzus persicae 13 3;9 95.5 87.4 Aphis fabae 11 2.5 86.1 61.5 Acyrthosiphon pisum 12 3.6 95.8 83.0 Megoura viciae 10 3.6 29.0 16.4

TABLZ 16 - Feeding behaviour of 4th. instar A. bipunctata. As with the lst.instar the larvae spent a similar time feeding on and ate similer proportions of M. persicae 113.

and A. pisum. Larvae feeding on A. fabae, however, took about five times as long, and, as in previous experiments, less of the prey was eaten. 4th. instar larvae of A. bipunctata capturing M. viciae accepted it readily as food. After about 4 mina. feeding, however, the predator suddenly rejected its prey and vomited, suggesting that this time was needed for the toxin to act. The figures given for - percentage weight of prey eaten are in this case unreliable, as they do not allow for food vomited by the predator. In spite of this behaviour repeated attempts to feed were made by larvae left in contact with M. viciae, and eventually a fatal dose was accumulated ( see pages 127-131 ).

Coccinella 7-punctata - 1st. instar (Table 17) 1st. instar larvae of C. 7-punctata, slightly larger than the 1st. instar of A. bipunctata, are rather more efficient at capturing their prey. Most times the first aphid encountered vas successfully captured. Again this was frequently accomplished in the long-legged species by first gripping hold of a tarsus, and only later, when the prey had tired through dragging the 114.

predator around, by transferring the mouthparts to the body of the aphid.

Mean time Mern no. ?lean time % wet % dry PREY No. of before 1st of prey spent weight weight reps. capture encountered feeding eaten eaten (mins) (mine) Myzus persicae 10 1.4 1.1 101 . 3 61.5 52.3 Aphis fabae 10 2.3 1.1 233.9 63.8 46.1 Acyrthosiphon pisum 10 2.3 1.1 82.4 90.8 80.7 Megoura viciae 10 3.8 2.5 422.8 58.6 33.8

TABLE 17 - Feeding behaviour of 1st. instar C. 7-punctata.

The time spent feeding on M. persicae and A. pisum was similar, and also comparable with the time taken by 1st. instar A. bipunctata on these two species. As also in A bipunctata the larvae took much longer to feed on A. fabae, but more of the aphid was eaten by C. 7- punctata than by A. bipunctata. Similar amounts (62-64%) of M. persicae and A. fabae were eaten, but the 1st. instar C. 7-punctata ate much more of A. pisum (91%).

115.

A. pisum seems to be generally more suitable for 1st. instar C, 7—punctatc, Larvae feeding on M. viciae took over five times as long as those feeding on A. pisum and consumed much less of the prey.

Coccinella 7-punctata - 4thoinstar (Table 18) 4th. instar larvae of C. 7-punctata appear to be very efficient predators.

Mean time Merl). no. Mean time % wet % dry PREY No. of before 1st of prey spent weight weight reps. capture encountered feeding eaten eaten (mins) (mins)

IiIyzus persice 10 1.6 3.3 100.0 100.0

Aphis 3.2 fabae 9 1.2 100.0 100.0 Acyrthosiphon pisum 11 1.6 5.0 96.4 93.6 Megoura viciae 12 0.5 4.6 94.2 90.3

TABLE 18 - Feeding behaviour of 4th. instar C. 7-punctata.

They quickly captured and ate all the species of aphid. Larvae feeding on M. persicae and A. fabae all 116.

consumed their aphids completely. When the prey was A, Disum one or two legs were invart-bly left behind, and when the prey was M. viciae most of the cuticle was usually left. Fig. 8 provides a comparison of some of the results just described. In order to get a ruantity directly proportional to the suitability of an aphid as food, the values for time spent feeding have been transformed into 'feeding rates', expressed as the reciprocal of the time spent feeding in minutes, multiplied by 100. The following points may be made from this comparison: (1) C. 7-punctata is a more efficient predator than A. bipunctata on the species of aphid used in the experi- ment, perhaps because it is larger. (2) The 4th. instar larva of both species is a far more efficient predator than the 1st. inc-tar, but then the relative size of the predator and its prey must be taken into account. The aphids presented to 1st. instar larvae were relatively much larger than those presented to 4th. instar larvae. The 1st. instar larvae might have fared better on aphids weighing less than 0.3 mg. Nevertheless observations of the behaviour of lst. instar

Mp Ap Af /Ay Mp Ap Af My

ADALIA BIPUNCTATA COCCI N ELL A 7- PUNCTATA

so

,c; 6° g y p

th

st

Mp Ap Af My Mp Ap Af My

ADALIA BIPUNCTATA COCC IN ELLA 7- PUNCTATA

FIGURE 8. Comparison of the Feeding Behaviour of 1st and 4th Instar Larvae on Four Aphid Species 118.

Coccinellid larvae under natural c)neitions suggest that very small aphids are not nece.sarily selected as prey. Of five 1st. instar larvae of C. 7-punctata released onto bean plants infested rith A. pisum, the first aphids caught by two were very large (3rd. or 4th. instar nymphs), two caught aphids of about the same size (2nd. instar) as those used in the feeding experi- ments, and one caught a very small (1st. instar) nymph. (3) Where a larva takes longer th:'n normal over feeding on an aphid, it generally consumes relatively less of the aphid. (4) The effect of an unsuitable food on the 1st. instar larva is more pronounced than on the 4th. instar. Table 19 makes a comparison of the effects of Aphis fabae as food for A. bipunctata and C. 7-punctata.

119.

Mean time Mean no. Mean time % wet % dry A. let No. of before 1st of prey spent weight weight INSTAR reps capture encounter ed feeding eaten eaten (mins) (mins)

Adalia bipunctata 13 2.6 1.5 379.0 30.4 12.7 Coccinella 2.3 233.9 63.8 46.1 7-pi nctata 10 1.1

4th No. of Mean time Mean no. Mean time wet % dry B. before 1st of prey spent weight weight INSTAR.reps capture encountered feeding eaten eaten (mins) (mins)

Adalia 43.4 86.1 61.5 bipunctata 11 1.7 2.5

Coccinella 10 1.2 1.0 3.2 100.0 100.0 7-punctata

TABLE 19 - Comparison of larvae of A. bipunctata and C. 7-punctata fed on A. fabae. 120.

SELECTION BEHAVIOUR BY THE LARVAL PREDATOR.

Whether Coccinellid larvae shovr any preference when given a choice of prey is a somewhat academic question, as larvae can rarely have a choice of species of aphid in the field. It seems nrobable that in most cases a larva completes its development on a single aphid species, and that the aphid food of the larva is determined by the site chosen by the parent to lay her eggs. Nevertheless, it is important to find out if the larval predator has any ability to select the most suitable aphid when given a choice, even if such an ability would normally be latent under natural conditions.

A. Adalia bipunctata

An experiment was conducted to see if Aphis fabae was recognised by larvae of A. bipunctata as an inferior food. Young larvae on the point -)f dispersion from the egg cluster were placed in individual rearing cells, and equal numbers of small nymphs of A. fabae and M. persicae were counted into each cell. The cells were examined daily and the number of live aphids of each species remaining were counted. Then the old aphid

121.

material was removed and fresh aphids, equal numbers of each species, were presented to each predator. As the larvae developed larger numbers of larger-sized aphids were used. The results are illustrflted in Pig. 9. The mean development time of A. bipunctata reared on a mixture of equal numbers of M. persicae an0 A. fabae was 12.2 days, mid-way between the development times of larvae reared on each aphid separately (Table 1). Until the 4th . instar (about the eighth day after hatching) no selection was shown, but the 4th. inptar exhibited a slight but definite preference for the more suitable aphid, M. persicae. The percentage of M. persicae in the total number of aphids t'c:en during development was in consequence significantly greater than the percentage of A. fabae. (Table 20).

MEAN TOTAL YUMB7R OF :EACH % OF EACH APHID SPECIES TY7IT SPECIES TAYEN SPECIES DAYS 1-8 DAYS 9-12 DAYS 1-12

Myzus 223.3 persicae 78.0 145.3 54.5 Aphis 74.7 110.5 185.2 45.5 fabae

TABLE 20 - Numbers of each species taken by A. bipunctata larvae given a mixture of equal numbers of Myzus persicae and Aphis fabae. 0 2 30 re O. 61J TOTAL N OF CH ES

FIGURE 9 MEAN O. EA SPECI TAKEN 120 40 20 160 80 40 240 50 10 Selection BehaviourbyLarvaeofAdaliabipunctata L. 2 2

NUMBER OFDAYSAFTERHATCHING NUMBER OFDAYSAFTERHATCHING 4 4

6 6

8 8

10 10 /

/

12 12 Myzus persicae Aphis faboe 123.

The results of this experiment cannot be said to show that 4th. instar larvae of A. binunctata are able to detect a less suitable aphid and to some extent avoid it. The small difference in the numbers of each species takalcould be due to a slight difference in the behaviour or mobility of the two aphids in the rearing cells, resulting in one species being captured more frequently than the other. The behaviour of both species of aphid in the rearing cells appeared to be identical, however, and it seems unlikely that such a difference would affect the feeding of the 4th. instar only. It was of interest to find out 7-hether larvae could be conditioned to select one species of aphid from a mixture, by previous feeding on that species alone. Accordingly, two groups of larvae were reared to the 4th. instar, one on M. persicae and the other on A. fabae. After the 3rd./4th. instar ecdysis all larvae were presented with a mixture of equU numbers of M. persicae and A. fabae, and the numbers of each species taken were recorded as before for twi successive days. The results are shown in Table 21, along with a comparison of the results for 4th instar larvae obtained in the previous experiment. 1211..

TREATMENT 5 OP APHIDS EATEN WHICH WERE A. FABAE 1st DAY 2nd DAY 3rd DAY

Larvae reared to 4th instar on 50/50 mixture 40.3 43.2 44.3 Larvae reared to 4th instar on M. persicae 43.7 37.5 •••• Larvae reared to 4th instar on A. fabae 39.6 31.2

TABLE 21 - Effect of previous diet on selection behaviour of 4th. instar A. binunctata.

On the first day that a choice was given, the previous food of the larvae did not significantly effect the proportion of A. fabae eaten. The larvae showed no sign of having been conditioned. On the second day, however, there was a significant chrtnge in the selection behaviour. Larvae which up until the 4th. instar had been given M. persicae now took significantly more N. persicae .arid less A. fabae' than before; but so also did larvae roared on A. fabae alone. It appears that larvae fed on a diet consisting of a single species of aphid ucre better able to select than those fed on a mixture; but 125.

in one case they were selecting the aphid with which they were familiar, and the other case, the unfamiliar. It was unfortunate that the experiment could not be continued because the majority of the larvae pupated on the third day. An experiment was carried out to test the ability of larvae of A. bipunctata to avoid an aphid which is toxic to them, Megoura viciae. Five groups of 12 larvae were separated at the time of dispersion from the egg- cluster and placed in separate rearing cells. The following treatments were given: (1) Larvae of one group were starved. (2) Larvae of a second group were given a suitable aphid species as food. Acyrthosiphon_pisum was chosen as it is similar in size and behaviour to Megoura viciae. Larvae of the third group were given M. viciae only. Larvae were given a mixture of equal numbers of M. viciae and A. pisum. Larvae were given a mixture of M. viciae and A. pisum in the proportions of one of the toxic aphids to nine of the suitable. 126.

The experiment was examined daily and records were made of the number of predators surviving and the numbers of each species of aphid taken. The cells were cleaned out and fresh aphids were presented daily in the same proportions as before. The times of survival of larvae given these five treatments are illustrated in Fig.10. A third of the larvae fed on A. pisum alone died, but the rest pupated on the 10th and 11th days after dispersion. Larvae fed on M. viciae alone were all dead on the second day, quicker than those which were starved. The rapid effect of the toxin in M. viciae has already been described. Of the larvae of group (4), five reached the second instar and one attained the third instar, the latter individual eventually dying on the tenth day after the start of the experiment having consumed a total of 14 A. pisum and 11 M. viciae. Even in group (5) (9 A. pisum: 1 M. viciae), none of the larvae reached the pupal stage, although seven reached the 4th. instar and two survived for 16 days, The longer a larva survived, the less susceptible it was to M. viciae. As has already been described, FIGURE 10 Percentage of larvae surviving 30 50 I Toxicity ofMegouraviciceBuckttolarvaeAdalia bipunctata. , I , , 1 \ , ,

i

I

I t

I I

I

I

I

i 1 1

1 c;

\ \

- —•-• \

\ \

\ \

b \

\ \ '-

6 \

Days afterdispersion ---- 8

\ \ \ \ 10 ---4k

\ 12 N

is.. 14

.. 16 't /*pm vicaralone M.viciae/ A.pisum M.v itioe/A.pisum Acrifo larvae slaved

1 9 , hon 9 18 50/50 10/90 done 11 . 11,

128.

4th. instar larvae reject M. viciae suddenly after about four minutes feeding. The seven larvae surviving to 4th. instar in group (5) captured and fed on an average of 7.3 M. viciae (and 86.9 A. pisum) in the 4th. instar before they died. In spite of the results of this experiment, which show that A. bipunctata larvae cannot avoid feeding on M. viciae when this aphid is presented in a mixture with A. pisum, the possibility of selection cannot be ruled out. In particular larvae of group (4), given equal numbers of M. viciae and A. pisum, took on average many more of the suitable aphid. Tables 22 and 23 show the numbers of each species of prey taken and the number of predators dying in each instar for groups (4) and (5). 129.

PREDATOR 1st INSTAR 2nd INSTAR 3rd INSTAR 4th INSTAR NO. Numbers t9.ken. Numbers taken Numbers taken Numbers tkn. A.p M.v A.p M.v A.p M.v A.p M.v

1 5 1 D 2 2 1 D 3 4 1 D 4 8 4 D 5 1 1 D 6 2 4 1 D 7 2 1 D 8 2 9 1 D 9 11 4 D 10 6 1 6 1 D 11 1 1 D 12 1 6 2 8 8 D

1 TOTAL 44 15 7 29 6 4 8 8 IMO died died died

TABLE 22 - Group (4): Larvae of A. bipunctata given equal numbers of A. pisum and M. viciae (D = died).

130.

- ), gild INSTLR 4th INSTAR PR - 1 t INSTAL 3rd 77STA:al TDATOR Numbers- Numbers Numbers Numbers NO. taken taken taIzer i taken A.p M.v A.p Mov A.p A.p rd.v

1 5 5 10 1 D 2 3 7 1 D 3 1 MO. 18 2 122 18 D

2 MED 16 2 148 16 D 5 2 7 9 87 9 D 6 2 1 D

7 1 ••• 8 49 9 D 8 2 9 3 1 D 10 1 4 9 2 115 11 D 11 1 4 9 50 5 D 12 1 5 VON• 10 001 37 4

1 TOTAL 24 1 1 608 72 7 3 died 44 died 89 I died died

TABLE 23 - Group (5): Larvae of A. bipunctata given A. pisum and Tic viciae in the Proportions 9/1. 131.

It was noticeable that although more than half the larvae in group (5) attained the 4th. instar, none pupated. Two groups of 12 larvae were reared to 4th. instar on A. pisum. They were given one day of feeding on A. pisum in the 4th, instar, and then one group was fed on M. viciae alone, while the other was starved. The larvae fed on M. viciae survived from 23-191 hours, with a mean survival time of 100.4 hours. None reached the pupal stage. Of the 12 starved larvae, however, seven pupated in the period 71-120 hours after the start of the experiment (mean time to pupation 85.6 hours), while the other five survived longer than the larvae fed on M. viciae (167-240 hours, mean survival time 196.2 hours). It would seem that feeding on M. viciae inhibits pupation.

B. Coccinella 7-punctata.

An experiment was done to see whether larvae of C. 7-punctata show any preference when given a mixture of equal numbers of A. pisum and M. viciae. The numbers of each species taken were recorded and fresh aphids were presented daily. As can be seen in Fig. 11, up to the 4th. instar (5-6 days after hatching) there was no

M. viciae

120 A. pisum

ILA

U u) X 80 1.1J 6 z0

72 40

2 6 8

M. viciae

A. pisum

NUMBER OF DAYS AFTER HATCHING

FIGURE 11

Selection Behaviour by Larvae of Coccinella 7- punctata 133.

definite preference, but in the 4th. instar there was a tendency to select the less suitable aphid (I. viciae). As in the experiments with A. bipunctata, a possible cause for this may be a slight difference in the behaviour of the two species of aphid in the rearing cells. In an attempt to overcome this difficulty, observ- ations were made of the behaviour of C. 7-punctata larvae presented with immobilised prey. An arena consisting of the base of a 3 in. petri-dish, with the sides coated with 'Fluon', was used. Aphids were carefully cemented down by their legs, in a 2 in. diameter circle in the positions of the numbers of a clock face. Four species were used, Acyrthosiphon pisum, Myzus persicae, Aphis fabae and Megoura viciae. Three individuals of each species, all about the same size, were selected, and cemented in alternate positions around the circle. The arena was illuminated from above. The predator was placed in the centre of the arena, and observed until it started feeding on an aphid. If it left the circle without encountering an aphid 134.

it was replaced in the centre. An encounter followed by feeding for 10 minutes by a 1st. instar predator, or for 2 minutes by a 4th. instar, as counted as an acceptance. If the prey vas touched with the mouthparts but abandoned before the critical time it uas counted as a rejection. The results are shown in Tables 24 and 25.

SPECIES OF NUMBER OF NUMBER OF NUMBER OF APHID ENCOUNTERS ACCEPTANCES REJECTIONS

Myzus persicae 28 21 7 Aphis fabae 28 23 5 Acyrthosiphon 2 pisum 28 26 Megoura viciae 25 22 3

TABLE 24 - Behaviour of 1st. instar C. 7-punctata with immobilised prey. 135.

SPECIES OF NUMBER OF NUMBER OF NUMBER OF APHID ENCOUNTERS ACCEPTANCES REJECTIONS

Myzus persicae 30 19 11 Aphis fabae 36 32 4 Acyrthosiphon pisum 37 30 7 Megoura viciae 39 29 10

TABLE 25 - Behaviour of 4th. instar C. 7-punctata with immobilised prey.

No definite preference or dislike was exhibited for any one of the four aphid species. Most of the aphids encountered were accepted. 136

SELECTION BEHAVIOUR BY THE ADULT PREDATOR

Adult Coccinellids are very active in the field and have a greater chance of selecting what they eat than do their larvae. The question of whether the adult shows any preference when given a choice of food is therefore of greater significance than for the larva, especially when it is considered that the ovi- position site and hence the food of the larva are likely to be related directly to a food supply chosen by the adult. In one experiment, 24 adults of A. bipunctata were confined singly in cells and presented with equal numbers of large nymphs of M. persicae and A. fabae. The adults had been reared as larvae on Acyrthosiphon pisum and were used only on their second and third days after emergence, no previous food having been given to them as adults. After the first 24 hours of the experiment the numbers of aphids of each species left were countermand the cells were cleaned out and fresh aphids presented to the predators. After a further 24 hour period the number of aphids left were 137.

again counted. The results are shown in Table 26, in terms of the percentage of each species taken.

PERCENTAGE OF EACH SPECIES TAKEN MYZUS PERSICAE APHIS FABAE S.D.

2nd DAY AFTER 69.8 30.2 ± 14.1 EMERGENCE

3rd DAY AFTER 65.2 34.8 ± 16.0 EMERGENCE

TABLE 26 - Selection by adult A. bipunctata given a mixture of equal numbers of Myzus persicae and Aphis fabae.

Thus, under the conditions of the experiment, adults of A. bipunctata showed a definite preference for the more suitable aphid, M. persicae. The difference between proportions eaten on the second and third days after emergence was not significant. A similar experiment was done using both A. bipunctata and C. 7-punctata and mixtures of A. pisum and M. viciae. Owing to a shortage of M. viciae the numbers of each 138.

species presented were not equal (i.e. 2 A. pisum: 1 M. viciae on the second day after emergence and 3 A. pisum: 1 M. viciae on the third day), so the results in Tables 27 and 28 show the percentages taken of the numbers of each species presented. In this experiment the adults were from larvae reared on M. persicae.

% TAKEN OF THOSE PRESENTED: A. PISUM M. VICIAE

2nd DAY AFTER 15.6 ± 9.5 9.2 ± 10.2 No significant EMERGENCE difference between the 3rd DAY AFTER two species EMERGENCE 29.5 ± 5.7 33.3 ± 14.8

TABLE 27 - Selection behaviour by adult A. bipunctata given a mixture of equal numbers of A. pisum and M. viciae.

Adults of A. bipunctata were not able to distinguish between the toxic aphid and the suitable one (Table 27). 139.

% TAKEN OF THOSE PRESENTED: A. PISUM M. VICIAE

2nd DAY AFTER 7.3 ± 5.4 31.7 ± 14.4 Significant EMERGENCE difference 3rd DAY AFTER between the 41.7 ± 25.3 82.5 ±24.0 EMERGENCE two species

TABLE 28 - Selection behaviour by adult C. 7-punctata given a mixture of equal numbers of A. pisum and M. viciae.

Adults of C. 7-punctata, however, were seemingly able to distinguish between the two species of aphid, but they apparently preferred the less suitable species, M. viciae. This recalls the preference of the 4th. instar larva of C. 7-punctata for M. viciae. Since A. bipunctata did not tend to select M. viciae under the conditions of the experiment, it seems unlikely that a difference in behaviour between A. pisum and M. viciae could have affected the results with C. 7-punctata, either of the selection experiments with larvae or of those with adults. If there had been such a difference it should have produced the same 140.

results with A. bipunctata as with C. 7-punctata. The work by Iperti (1965) referred to in the previous section has shown that habitat factors rather than food preferences are likely to play a predominant part in oviposition-site selection by adult Coccinellids. In an experiment to study this aspect of oviposition behaviour, adults of A. bipunctata and C. 7-punctata were given a choice of oviposition sites at different height levels. Field cages 5ft. high x 4 ft. x 3 ft. were used. Broad bean plants in pots were suspended by means of wires at heights of 4ft. and 2ft. 6in., four pots at each height, and a further four pots were placed on the gravel floor of the cage (Fig. 12). The plants were infested with Acyrthosiphon pisum, and renewed every two weeks. In August 1964, 15 pairs of young adults of each species of Coccinellid were introduced into the same field cage. The numbers of eggs laid on the plants at each level were recorded on alternate days. In practice it was found difficult to tell the eggs of the two species apart, so they were hatched out in the laboratory to check identification. In July/August 1965 the experiment was repeated using a separate field FIGURE 1 2

Field cage used in experiments on oviposition site selection by adult Coccinellids.

143.

cage for each species of Coccinellid, and 20 pairs of each species were used. The results for both years are shown in Table 29. 1964

A. BIPUNCTATA C. 7 -PUNCTATA LEVEL Total Total Total Total Total Total no. of egg adults no. of egg adults eggs batches observed eggs batches observed HIGH 312 27 15 1 (4ft)

MEDIUM 123 11 13 5 (2-ift) LOW 113 8 10 89 8 15 (6in)

1965

A. BIPUNCTATA C. 7-PUNCTATA LEVEL Total Total Total Total Total Total no. of egg adults no. of egg adults eggs batches observed eggs batches observed

HIGH 426 32 17 ma. 1. (4ft) MEDIUM 140 13 7 11 (2ift) LOW 101 9 12 40 3 26 (6in)

TABLE 29 - Selection of oviposition site in relation to height in A. bipunctata and C. 7-punctata. X44.

In both years the experiment ended when no more eggs were laid, probably due to the shortening day length at the end of the summer. The numbers of eggs recorded for C. 7-punctata were very small; this species tended to lay its eggs on the mesh and the wooden framework of the cage rather than on the plants, and only eggs on the plants or pots were counted. Never- theless the results, agreeing closely for the two years, give a definite indication of a difference in oviposition behaviour between the two species. In agreement with the findings of Iperti (1965), A. bipunctata preferred to lay its eggs at a higher level than C. 7-punctata. Adults of A. bipunctata appeared to move freely from one height to another, in spite of their oviposition preference (Table 29). Adults of C. 7-punctata, however, tended to remain on the ground-level plants and ignore the other two layers, although they were often seen on the roof of the cage in sunny weather. The total numbers of adults observed on each layer were so small because most recordings were made in the evening, when the beetles had ceased to be active and were mainly to be found in the corners of the framework of the cage. 145.

EXAMINATION OF THE GUT CONTENTS OF FIELD-COLLECTED ADULTS

During the spring and summer of 1965 adults of Adalia bipunctata and Coccinella 7-punctata were collected from various localities in the field. The specimens were fixed in 75% alcohol, and the crop and intestine were dissected out, opened on a microscope slide with a drop of polyvinyl lactophenol, and covered with a cover-slip. The food in the gut was examined under a microscope.

ADALIA BIPUNCTATA 17 specimens examined; 6 males, 11 females. All collected at Silwood Park. 1 male; from Euonymus with Aphis fabae, 13/5 - few remains of black aphids in small pieces. 1 male; from groundsel (Senecio), 25/5, with several species of aphid in the vicinity - gut contained no aphids but numerous pollen grains, particularly of Pinus. 1 male, 4 females maturing eggs; from Sambucus with Aphis sambuci, 10/6 - all these specimens had gut practically empty. One female had 2 Pinus pollen grains. Possibly they had just arrived from another locality. 146.

1 female (maturing eggs); from field beans (Vicia faba) with Aphis fabae, 28/6 - gut empty. 1 male; from thistle (Cirsium) with Dactynotus sp., 28/6 - crop empty. Remnants of food in hind gut but these were unidentifiable. 2 males, 4 females (maturing eggs); potato field, 29-30/6 - all had only a small amount of food in the gut, mainly aphids. Two of the females had some pollen grains. 2 females; from field beans with Aphis fabae, 8/7 - one contained some black aphid remains and 10-12 pollen grains, the gut of the other was empty.

COCCINEI►I,A 7-PUNCTATA 23 specimens examined; 16 males, 7 females. All except the first two collected at Silwood Park. 2 females (without eggs); Scilly Isles, April - gut virtually empty. 6 males, 1 female maturing a few eggs; on ground, in newly ploughed area between two plots of Brassicas, 11/5 - all but 3 had no food. Female had very few aphid remains in small pieces. One male had remains of small aphids and a few pollen grains; 14.7.

another male had eaten quite large aphids and many pollen grains. 1 female (without eggs); from Sarothamnus, with no aphids, 13/5 - remains of one or more mites or small spiders and a few pollen grains. 1 male; from groundsel (Senecio) with several species of aphids in the vicinity, 25/5 - remains of several aphids, no pollen grains. 1 female (without eggs); potato field, 30/6 - many aphid remains. 1 male; from isolated raspberry shrub, 13/7 - remains of at least one aphid. 2 males; from thistle (Cirsium) with Aphis fabae, 22/7 - one had gut crammed with aphid remains. The gut of the other was empty. 2 males; from vetch (Vicia sativa) with Megoura viciae, 22/7 - both had remains of large aphids with pigmented tarsi, almost certainly Megoura. 1 male; from broom (Sarothamnus) with Aphis sarothamni, 28/7 - gut full of remains of black aphids. 1 male, 1 female; from thistle with Dactynotus sp., 3/8 - male had gut crammed with aphid remains, including large aphids which could be 148.

Dactynotus. Female had a few large aphids. 2 males, 1 female (without eggs); from thistle with Aphis, fabae, 3/8 - one male had gut full of black aphids, the other had no food. Female had remains of one or two aphids, but gut was almost empty.

This work was in the nature of a preliminary invest- igation only, but nevertheless produced some interesting results. The number of adults collected which had little or no food in their gut, even at a time when aphids of many species were abundant at Silwood Park, was surprising, When aphid food was scarce early in the season, both Coccinellids were found with pollen grains. None of the Adalia bipunctata collected from Sambucus with Aphis sambuci had fed on this aphid. Possibly this was a coincidence, but it does illustrate the danger of assuming feeding without actually observing it. There is no doubt that A. bipunctata frequently lays its eggs and develops its larvae on Aphis sambuci, so it would be interesting if the adults did not feed on this species in the field. Coccinella 7-punctata was collected from vetch (Vicia sativa) with Megoura viciae, and the food in the gut 149. appeared to be M. viciae. One individual was observed feeding on M. viciae in the field. This aphid increased mortality and appeared to inhibit egg production of adult C. 7-punctata in the laboratory (p.96 ), yet seemed to be preferred to Acyrthosiphon pisum (p.139). The habitats of C. 7-punctata and Megoura viciae coincide and it seems probable that they frequently meet under natural conditions. The larvae are able to develop on M. viciae, so it seems that the larvae have become adapted to feeding on this aphid while the adults are still adversely affected. It is clear that a more extensive study of the gut contents of aphidophagous Coccinellidae would provide valuable information on the food utilised by different species in the field, which could not be reliably deduced from distribution studies, and which would parallel laboratory investigations into the suitability of differ- ent foods. 150.

DISCUSSION

The results of the experimental work on Adalia bipunctata and Coccinella 7-punctata permit the following conclusions concerning these two species;

1. There can be considerable variation in suitability of different species of aphid which are accepted as prey. Certain aphid species, for example Myzus persicae, Aulacorthum circumflexum, Acyrthosiphon pisum and Microlophium evansi in the case of Adalia bipunctata, seem to satisfy all the nutritional requirements for development of the larvae. Other aphids, for example Aphis fabae, Aphis sambuci and Brevicoryne brassicae, are relatively unsuitable, and some such as Megoura viciae are toxic. The effects of different aphid species on adult fecundity may vary in the same way.

2. The two Coccinellid species differ in their range of suitable prey. For example, Megoura viciae is toxic to Adalia bipunctata but only slightly unsuitable for larval development of C. 7-punctata; Aphis fabae is relatively unsuitable as food for larvae and adults of Adalia bipunctata but seems to satisfy all the nutritional 151.

requirements of C. 7-punctata. From a comparison of the effects of only a small number of aphid species it is evident that these two Coccinellids have a considerable degree of physiological specificity.

3. The Coccinellids apparently have no ability to detect and avoid unsuitable or toxic species. The experiments on prey selection sometimes indicated slight preferences, but the preference was not always for the more suitable food. C. 7-punctata, for instance, appeared to prefer Megoura viciae to Acyrthosiphon pisum. Although larvae of Adalia bipunctata seemed to prefer Acyrthosiphon pisum when given this aphid mixed with Megoura viciae, they could not avoid feeding on the toxic aphid and eventually died.

The results of laboratory choice experiments must of course be interpreted with extreme caution. It could be argued that under laboratory conditions a Coco inellid provided with an excess of food is most likely to demonstrate any preference it may have for one prey over another. Therefore laboratory work would tend to emphasise any ability to select, and probably only marked preference for one aphid or a strong 1 52.

avoidance reaction to another would have any significance when applied to conditions in the field. On the other hand, a Ooccinellid has more chance of avoiding an unsuitable aphid in the field, than in the confined conditions of a laboratory experiment. The adult beetle finding an aphid colony which is an unsatisfactory source of food may consume a few aphids and then become restless and fly away, with a good chance of subsequently encountering a more suitable species on which it will stay longer and possibly reproduce. Adults of Adalia bipunctata, which in the laboratory seemed unable to distinguish Megoura viciae from Acyrthosiphon pisume may in this way avoid prolonged contact with Megoura viciae in the field. Owing to its preference for higher levels of vegetation, Adalia bipunctata may not often encounter Megoura viciae, but this sometimes occurs as in July 1965 at Silwood Park. when rows of field beans infested with Megoura viciae were adjacent to rows infested with Aphis fabae. Numerous adults and larvae of Adalia bipunctata were found on the Aphis fabae-infested plants, but none in the vicinity of the Megoura viciae.

4. Aphis fabae and Aphis sambuci were found to be 153,

relatively unsuitable food for Adalia bipunctata, despite the fact that they are both common prey of this Coccinellid under natural conditions. The increased development time and increased mortality may be relatively unimportant to the species under natural conditions, especially as the food of the larva has apparently no effect on the fecundity of the resulting adult; but the reduced fecundity of adults feeding on Aphis fabae is particularly significant. Aphis fabae is an "essential" food of Adalia bipunctata, according to Hodek's terminology, but it is an essential food which does not allow the Coccinellid to realise its full reproductive potential.

Reasons for the unsuitability or toxicity of certain aphid species are difficult to determine. The unsuitability of Aphis fabae to larvae of Adalia bipunctdta has already been discussed. In this case two factors appear to be involved, difficulty in ingesting the food once the prey is captured, and low nutritive value. The two may be related; the nutritive value may be low because some essential nutrient is being left behind in that part of the prey which is not ingested. In other 154

instances a third factor, difficulty in capturing the prey (eg. A.F.G. Dixon 1958), may play a part. This is least likely for adults and fourth instar larvae, which are generally relatively efficient at capturing and eating prey; for them the nutritive value of the food is likely to be the most important factor influencing suitability. Difficulty in catching and eating the ftod may be relatively more important for younger larvae. The aphid host plant may affect the value of the aphid as food. However, Aphis fabae had the same effect on larvae of Adalia bipunctata whether it was obtained from Vicia or Euonymus and Myzus persicae reared on Brassica and on Vicia were equally suitable as food for Coccinella 7-punctata larvae. Aphids which apparently contain toxic substances may obtain them from the host plant. Hodek (1956, 1957) suggests that the adverse effect of Aphis sambuci on Coccinella 7- punctata may be due to the presence of the glycoside sambunigrin obtained by the aphids from the elder (Sambucus nigra), which could be broken down by the Coccinellid to give hydrocyanic acid. The toxicity of Megoura viciae to Adalia bipunctata (and to Adalia 155.

10-punctata, Dixon 1958) may also be due to some substance obtained directly from the host plant, which is not metabolised by the aphid. Investigations by Dixon et al.(1965) have not yet revealed the nature of the toxin in Megoura viciae. If we now incorporate the habitat selection behaviour shown by the adults, we begin to appreciate the various interacting aspects of specificity shown by Adalia bipunctata and Coccinella 7-punctata. The adult Coccinollids are able to disperse widely and search over large areas for their food. Their distri- bution will depend primarily on the location and abundance of food. If there is plenty of food at all levels of vegetation, then the two species will show a tendency to feed, and especially to lay their eggs, at different levels. The preference seems to be oriented towards a particular height above ground rather than towards a particular type of vegetation or microclimate; it is still in evidence when similar oviposition sites are presented at different heights in a field cage. Thus Coccinella 7-punctata prefers to be at ground level, and is found most commonly in low 156

field crops or meadow land, while Adalia bipunctata prefers, at least for oviposition, vegetation 4 feet or more above the ground, and therefore occurs most commonly on aphid infestations on trees and shrubs, as well as on tall crop plants such as field beans or maize. Within their preferred habitats the two species will encounter aphids which vary in value as food. Some species of aphids will satisfy all their nutritional requirements, some may oe harmful, others may be harmless but will not promote reproduction and larval development, and others may permit reproduction and development but still be relatively unsuitable. It seems unlikely that these two Coccinellids actively avoid unsuitable aphids. It may happen, however, that an adult feeding on aphids which fail to provide a satisfactory food will soon become restless and leave. Alternatively if it remains it may die or fail to mature its eggs, depending on the nature of the unsuitability of the aphid. Either way it is possible that eggs are rarely laid in the vicinity of a food which will not permit development of the larvae. However we have seen that a relatively 157.

unsuitable species of aphid can be common prey of both adults and larvae. :At. times when aphid food is scarce, the Coccinellids will turn to other food sources. Some of the foods utilised in an emergency are described in a previous section. Probably their increased restlessness brings the Coccinellids into contact with alternative foods which they would not normally_ encounter, including other insects, mites, pollen and fungal spores. Aphidophagous Coccinellids in general may .cossess this subtle form of specificity, whereby they can. probably utilise to their best advantage. the largely unpredictable outbreaks of aphids as they arise, laying their eggs and developing their. larvae when suitable aphid food is plentiful. The differences in choice of habitat and range of suitable prey will both help to reduce interspecific competition. When suitable aphid food cannot be found they can survive on alternative, emergency foods. It is surprising that in view of the outstanding successes obtained with predators in early biological control work, the potential value of predators as 158.

biological control agents should be a matter of controversy; yet it is a widely held view that in most cases parasites are more likely to succeed than predators. The reputation of predators has suffered on two accounts;

1. Predators are still considered to be generally non-specific, while parasites are regarded as specific. This conception has been attacked by Thompson (1928, 1951) and by Chung-lo (1958) but it still remains in the minds of many workers, despite Chung-lo's statement that "it may be assumed that this rather general misconception about the predator has by now been finally dispelled". The number of predators which can be called truly non-specific diminishes as their biologies become better known, as also does the number of parasites which can be credited with strict monophagy.

2. There is still a tendency to regard specificity as an altogether beneficial character when assessing the potential value of an entomophagous insect as a control agent. This belief probably stems from the theories of "density dependence" established by Lotka, Volterra and Nicholson, which deal only with specific predators 159. •

and parasites. Also, some of the classic successes in biological control have been obtained with parasites and predators commonly regarded as highly specific, although Chung-lo (1958) points out that very few of these successful biological control agents are in fact specific in the strictest sense. In an important paper, Mesnil (1958) concludes that the value of non-specific predators and parasites should not be underestimated, and that they should be studied more intensively. The advent of the integrated control concept has given new significance to the importance of non-specific parasites and predators. Several workers, eg. De Bach. (1951) and Pickett (1959) have recognised the important part that can be played by so-called general predators and other native natural enemies if they are encouraged in integrated control programmes. The ecological approach to pest control involves far more than the culture and release of a specific parasite or predator to combat a specific pest, and it is necessary that in the future the far more numerous, relatively non-specific parasites and predators are increasingly utilised. In order to use them effectively we need to study every 160.

aspect of their ecology and to gain an intimate knowledge of their relations with other species so that the exact part which they can play in a general scheme of pest control can be determined. We know that with aphidophagous Coccinellidae both nutritional and ecological factors must be taken into consideration in assessing whether a particular predator is likely to be useful against a particular pest. We also know that alternative foods are essential and that it is necessary to study these foods on wild plants, in both wild and cultivated habitats, because it is at times when it is away from the crop that the predator needs to be encouraged, in order that adequate numbers of the right species will emigrate to the crop at the right times of the year. 161.

SUMMARY

1. Larvae of Adalia bipunctata developed equally well on several aphid species; Mvzus persicae, Aulacorthum circumflexum, Acyrthosiphon pisum, and Microlophium evansi. Aphis fabae and Aphis sambuci were relatively unsuitable prey, resulting in increased development time, greater mortality and smaller adults. Aphis fabae was equally inferior whether from Vicia or from Zuonymus. The total fresh weight of Aphis fabae consumed during larval development was similar to that of suitable prey, but the rate of uptake of food was slower. Brevicoryne brassicae was also unsuitable, and Megoura viciae was toxic to A. bipunctata larvae.

2. The development of larvae of Coccinella 7-punctata on A. fabae was as fast as with any other aphid species. Larvae of this Coccinellid were also able to develop on Megoura viciae, but at a reduced rate. Aphis sambuci and Brevicoryne brassicae were unsuitable foods for C. 7- punctata larvae. Myzus persicae was equally suitable for larval development whether reared on Vicia or on Brassicae.

3. Aphis fabae was also relatively unsuitable for adult Adalia bipunctata in the adult stage. Fecundity was more 162.

than halved in comparison with adults fed on I4yzus

persicae; the size of eggs and the fertility were also reduced. The food of the larvae seemed to have no effect

on the fecundity of the resulting adult. In C. 7-punctata

feeding on A. fabae did not result in reduced fecundity.

Megoura viciae was toxic to adult A. bipunctata; C. 77

Punctata could survive feeding on this aphid but laid no

eggs.

4. A quantitative comparison of Aphis fabae and Myzus persicae as food for 4th. instar larvae of A. bipunctata

revealed that larvae feeding on A. fabae consumed relatively

less of each aphid which they captured, and that the material

which was ingested was apparently less nutritious than that

obtained from Hi persicae.

5. C. 7-punctata is a more efficient predator than A. bipunctata, and the 4th. instar larvae of both species

are far more efficient at capturing and eating prey than

the lot. instar larvae. The effect of an unsuitable food

on the 1st. instar was more pronounced than on the 4th.

instar. Where a larva took longer than normal over feeding on an aphid, it generally consumed less of the aphid.

6. A. bipunctata larvae fed a mixture of A. fabae and 163.

Myzus persicae showed a slight preference for M. persicae when they reached the 4th. instar. The preference for M. persicae seemed to be stronger when larvae were reared on a single species of aphid, whether A. fabae or M. persicae, up to the 4th. instar, and only then given a choice. Larvae of A. bipunctata could not avoid feeding on M. viciae when this aphid was presented in a mixture with Acyrthosiphon pisum, with fatal results; even when larvae were given a mixture containing one M. viciae in every ten aphids none survived to pupate.

7. Larvae of C. 7-punctata showed a tendency to select the less suitable M. viciae from a mixture of equal numbers of this aphid and A. pisum. In experiments in which larvae were presented with immobilised prey, however, no definite preference or dislike for any one aphid species was observed:

8. Adults of A. bipunctata showed a definite preference for Myzus persicae over Aphis fabae, but were unable to distinguish between Acyrthosiphon pisum and Megoura viciae. Adults of C. 7-punctata, however, were seemingly able to distinguish between A. pisum and M. viciae, but, like their larvae, preferred the less suitable aphid.

164.

9 When adults were given a choice of oviposition sites consisting of broad bean plants in pots suspended at three different heights and infested with A. pisum, there was a distinct difference in oviposition behaviour between the two species. A. bipunctata preferred to lay its eggs at the highest level (4 ft.), whereas C. 7-punctata tended to feed and oviposit on the ground-level plants.

10. These results provide further illustration of some of the factors involved in the food specificity of Coccinellidae, which have only become apparent in recent years. Knowledge of these factors is essential if we are to assess the part which can be played by such predators in the control of insect pests. 165.

A CKNOITLEDG=NTS

I wish to thank Mr. M.J.Way for his valuable advice and encouragement during the course of the work and for his criticisms of the manuscript. I am also grateful to Professor 0.W.Richards for granting me facilities in his Department at the Imperial College Field Station, Silwood Park, Ascot. The work was done while in receipt of a Department of Scientific and Industrial Research (Science Rosearch Council) Research Studentship. 166.

BIBLIOGRAPHY

ANDERSON, N.H. (1962a) Bionomics of six species of Anthocoris (Heteroptera:Anthocoridae) in Lngland. Trans. R. Ent. Soc. Lond. 114, 67-95. ANDERSON, N.H. (1962b) Growth and Fecundity of Anthocoris spp. reared on various prey (Heteroptera:Anthocoridae). Ent. exp. & Appl. .2, 40-52. BALDUF, W.V. (1935) The Bionomics of Entomophagous Coleo- ptera. St. Louis 220pp. BALDUF, 1Y.V. (1938) The Rise of Entomophagy among Lepido- ptera. Amer. Nat. /a, 358-379. BALDUF, W.V. (1939) The Bionomics of Entomophagous Insects. Vol. II. New York 384pp. BALDUF, W.V. (1942) Interrelations of Insect-eating Insects. Trans. Ill. Acad. Sci. Springfield. 21, 224-230. BANKS, C.J. (1954) Random and Non-Random Distributions of Coccinellidae. J. Soc. British Ent. 4, 211-215. BANKS, C.J. (1955) An Ecological Study of Coccinellidae associated with Aphis fabae. Bull. ent. Res. 46, 561-587. BANKS, C4J. (1956a) Observations on the behaviour and 167.

mortality in Coccinellidae before dispersal from the egg shells. Proc. R. ent. Soc. Lond. A 11, 56-60. BANKS, C.J. (1956b) Distribution of Coccinellid egg- batches and larvae in relation to numbers of Aphis fabae on Vicia faba. Bull. ent. Res. 4 2, 47-56. BARNES, H.P. (1929) Gall Midges as Enemies of Aphids. Bull. ent. Res. 20, 433-442. BARNES, H.F. (1930) Gall Midges as Enemies of Tingids, Aleyrodidae and Coccidae. Ibid. 21, 319-329. BARNES, H.F. (1933) Gall Midges as Enemies of Mites. Ibid. 24, 215-228. BENSON, R.B. (1950) An introduction to the natural history of British sauflies (Hymenoptera:Symphyta). Trans. Soc. British Ent. Bournemouth. 10, 45-142. BEQUAERT, J. (1922) The Predacious Enemies of Ants. Bull. American Mus. Nat. Hist. 20, 271-331. BISHARA, I. (1934) The Cotton Worm, Prodenia litura F., in Egypt. Soc. Roy. ent. d'Egypte Bull. 288-420. BOMBOSCH, S. (1962) Untersuchung uber die Auslosung der Eiablage bei Syrphus corollae F. (Diptera:Syrphidae). Z. Ang. ent. 50, 81-88. BRISTOWE, W.S. (1948) Notes on the habits and prey of 20 spp. 168.

of British hunting wasps. Proc. Linn. Soc. London 160, 12-37. BRUES, C.T. (1946) Insect Dietary. An Account of the Food Habits of Insects. Bayard U.P. 492pp. CAMPBELL, R.E. (1926) The Pea Aphis in California. J. Agric. Res. 22, 861-881. CHUMAKOVA, B.M. (1949) The effect of nutrition during the larval and adult stages on the fertility of predatory insects. (In Russian) Ent. Obozr. Moscow 0,, 225-234. CHUMAKOVA, B.M. (1962) Experiments in rearing a predatory beetle Cryptolaemus montrouzieri Muls. on an artificial diet. Zeszyty Prob. Postepow Nauk Roln. 11, 195-200. CHUNG-LO, LIU (1958) Monophagy versus Polyphagy in the Choice of Entomophagous Insects in Biological Control. Trans. Int. Conf. Ins. Path. Biol. Cont. Praha 521-531. CLARK, A.H. (1926) Carnivorous Butterflies. Smithson. Inst. Ann. Rpt. 1925, 439-508. CLAUSEN, C.P. (1916) Life History and Feeding Habits of a Series of California Coccinellidae. Calif. Univ. Pub. Ent. 1, 251-299. CLAUSEN, C.P. (1940) Entomophagous Insects. New York 688pp. CONRAD, M.S. (1959) The spotted lady beetle Coleotnegilla 169

maculate DeG., as a predator of European corn borer eggs. J. econ. Ent. 32, 843-847. CRISMAN, A.U. & DUMESTRE, J.0. (1930) The feeding rate of the Australian lady beetle Rodolia cardinalis. J. Agric. Res. 41, 197-203. DAVIES, M.J. (1953) The contents of the crops of some British Carabid beetles. Ent. mon. Mag. 12, 18-23. DAVIES, M.J. (1959) A contribution to the ecology of species of Notiophilus and allied genera (Coleoptera:Carabidae). Ent. mon. Meg. 21, 25-28. DeBLCH, P. (1951) The necessity for an ecological approach to pest control on citrus in California. J. econ. Ent. 44, 443-447. DEMPSTER, J.P. (1963) The natural prey of three species of Anthocoris (Heteroptera:Anthocoridae) living on broom (Sarothamnus scoparius L.). Ent. exp. & Appl. 6, 149-155 DIXON, J.F.G. (1958) The escape responses shown by certain aphids in the presence of the Coccinellid Adalia 10- punctata L. Trans. R. Ent. Soc. Lond. 110, 318-334. DIXON, A.F.G. (1959) An experimental study of the searching behaviour of the predatory Coccinellid beetle Adalia 10-punctata L. J. Anim. .col. 28, 259-281. DIXON, T.J. (1959) Studies on Oviposition Behaviour of 170.

Syrphidao (Diptera). Trans, R. Ent. Soc. Lond, 111, 57-80. DONISTHORPE, H.St.J. (1913) The Coleoptera of the British Isles. Fowler and Donisthorpe, London. 6, 201. EDWARDS, W.H. (1886) On the Life History and Preparatory Stages of Feniseca tarquinius F. Caned. Ent. 18, 141-153. EVANS, H.E. & YOSHINOTO, C.M. (1962) The ecology and nesting behaviour of the Pompilidae of the north-eastern United States. Misc. Publ. ent. Soc. Amer. Baltimore.,, 65-119. FORBES, S.A. (1880) Notes on Insectivorous Colooptera. Ill. State Lab. Nat. Hist. Bull. No.3. 1, 153-160. FORBES, S.A. (1883) Food Relations of Carabidae and Coccin- ellidae. Ibid. No.6. 1, 33-64. GRAND', G. (1959a) Contributi alla conoscenza degli Imenot- teri Aculeati. Indice analitico dei contributi 1(1925) XXVII(1957). Boll. 1st. Ent. Univ. Bologna z2, 69-85. GRANDI, G. (1959b) Contributi alla conoscenza degli Imenot- teri Aculeati XXVIII. Ibid. 12, 239-292. HAMM, A.H. & RICHARDS, 0.1T. (1926) The Biology of the British Crabronidae. Trans. R. Ent. Soc. Lond. 297-331. HAWKES, 0.A.M. (1920) Observations on the Life History, Biology and Genetics of the Ladybird Beetle, Adalia bi- Dunctata. Proc. zool. Soc. Lond. 1920, 475-490. 171.

HEPPER, H.C. (1959) Adalia bipunctata L. colaborador, en la lucha biologics. Lctas. Trab. I. Congr. sudamer Zool. 1959 La Plata (1960), 73-76. HILL, A.R. (1957) The biology of Anthocoris nemorum L. in Scotland (Hemiptera:Anthocoridae). Trans. R. Ent. Soc.

Lond. 109, 379-394. HILL, A.R. (1961) The biology of Anthocoris sarothamni Douglas and Scott in Scotland. Trans. R. Ent. Soc. Lond. 111, 41-54. HINGSTON, R.V.G. (1925) An oriental hunting wasp, Sphex lobatus. Bombay Nat. Hist. Soc. Jour. 22, 736-743. HINGSTON, R.V.G. (1926) An oriental hunting wasp, Sphex

lobatus II. Ibid.,, 147-159. HOBBY, B.M. (1931a) The prey of dungflies (Diptera:Coryluridae). Proc. ent.Soc. Lond. 6, 47-49. HOBBY, B.M. (1931b) The prey of Coenosia tigrina F. and other species of the genus. Proc. ent. Soc. Lond. 6, 13-15. HOBBY, B.N. (1932) The prey of sawflies (Hymenoptera: Tenthredinidae). Proc. ent. Soc. Lond. 2, 14-15. HOBBY, B.M. (1934) Notes on Predacious Anthomyidae and Cordyluridae (Diptera). Ent. mon: Mag. 22, 185-190. HOBBY, B.M. & SMITH, K.G.V. (1961) The bionomics of Empis opaca. Ent. mon. Mag. 22, 204-208. 172.

HODEK, I, (1956) The influence of Aphis sambuci L. as prey of the ladybird beetle Coccinella 7-punctata L. Acta Soc. Zool. Bohemoslov. 20, 62-74. HODEK, I. (1957) The influence of Aphis sambuci as food for Coccinella 7-punctata II. Acta Soc. Ent. Cechoslov. LA, 10-17. HODEK, I. (1959) Influence of various aphid species as food for Coccinella 7-punctata and Adana bipunctata. In The Ontogeny of Insects, Praha. 314-316. HODEK, I. (1962) Essential and Alternative Food in Insects. XI Inter. Cong. Ent. (Berlin) 1960 Bd.II 698-699. HODEK, I., STLRY, P. & STYS, P. (1962) The Natural Enemy Complex of Aphis fabae and its effectiveness in control. XI Inter. Cong. Ent. (Berlin) 1960 Bd.II 747-749, IBRAHIM, M.M. (1955) Studies on Coccinella 11-punctata aegyptiaca Reiche II. Biology and Life History. Bull. Soc. Ent. Egypte. 12, 395-423. IPERTI, G. (1961) Coccinellids. Their use in Agriculture. (In French) Rev. Zool. agric. 60, 14-30 and 60-71. IPERTI, G. (1965) Contribution a l'etude de la specificite chez les principales Coccinelles aphidiphages des Alpes -Maritimes et des Basses-Alpes. Entomophaga 10, 159-178. 1 73 .

JOHNSSEN, A. (1930) Beitrage zur Entwicklungsund Ernahr- ungsbiologie einheimischer Coccinelliden ureter besonderer Berucksichtigung von Coccinella 7-punctata. Z. Lug. Ent. 16, 87-158. KADDOU, I.It. (1960) Feeding Behaviour of Hippodamia quinque -signata Kby. larvae. Univ. Calif, Publ. Ent. 16, 181-232. KANERVO, V. (1940) Beobachtungen and Versuche zur Ermitt- lung der Nahrung einiger Coccinelliden. Lnn. ent. fern. Helsingfors 6, 89-110. KILLINGTON, J.F, (1932) Neuroptera feeding upon Lepidoptera. J. Ent. Soc. S. England. 1, 33-34.

KILLINGTON, J.F. (1936) L monograph of the British Neuro- ptera. London. Vol I. KOIDE, T. (1961) Observations on feeding habit in Coccinella 7-punctata bruckii Muls. (In Japanese) Kontyu Tokyo. 29., 211-213. KROMBEIN, K.V. (1958) Miscellaneous prey records of solitary wasps III (Hymenoptera: Lculeata). Proc. biol. Soc. liashington. a, 21-26. LINDROTH, C.H. (1949) Die fennoskandischen Carabidae III Stockholm, 1949. MESNIL, L.P. (1958) Considerations of the use of Polyvalent 174.

parasites or predators in biological control. Trans. Int. Conf. Insect Path. Biol. Cont. Praha 427-440. MILNE, D.L. (1960) The Gall Midges (Diptera:Cecidomyidae) of Clover Flower Heads. Trans. R. Ent. Soc. Lond. 112, 73-108. MISRA, M.P., NEGI, P.S. & GUPTA, S.N. (1930) The Noctuid I'Ioth (Eublemma amabilis); a Predator of the Lac Insect and its Control. J. Bombay Nat.Hist.Soc. 21, 431-446. MOTJRSI, L.L. & KAMM, M. (1946) Notes on the Biology and Feeding Habits of the introduced beneficial insect Leis conformis Boisd. Bull.Soc.Fouad Ier.Ent. 22, 63-74. MUIR, F. (1931) Introduction - in 'The Insects and other Invertebrates of Hawaiian Sugar-cane fields. by Francis Expt.Sta. Hawaiian Sugar Planters' J.ssocn. NIELSON, & CURRIE, W.E. (1960) Biology of the convergent lady beetle when fed a spotted alfalfa aphid diet. J. econ. Ent. .22, 257-259. NIJVELDT, W. (1954) The range of prey of Phenobremia urti- cariae Kffr, and Ph. aphidivora Rubs. (Diptera:Itonidae). Ent. Ber., Amsterdam. 11, 207-211 and 436-437. NINOMIYA, E. (1957) Food Habits of various aphidophagous Syrphid larvae. Jap. J. Lppl.Ent.Zool. 1, 186-192. (In Japanese; see Biol. Abstr. No.43271 (1959). 175.

OKAMOTO, H. (1961) Comparisons of ecological characters of the predatory ladybirds, Coccinella septempunctata bruckii fed on the apple grain aphids, Rhopalosiphum prunifoliae and the cabbage aphids, Brevicoryne brassicae. Jap. J. Lppl.Ent.Zool. 1, 277-278. In Japanese. OKUNO, T. (1961) Feeding tests and some field observations of three aphidophagous Coccinellids. Pub. ent. Lab., Univ. Osaka Pref, No.6, 149-152. PALMER, M.J. (1917) Additional notes on heredity and life history in the genus Adana. Ann.ent.Soc.Amer. 10, 289-302. PAR 3R, J.R. (1918) The Life History and Habits of Chloro- Pisca gabra Meig., a Predacious Oscinid. J. econ. Ent.

11, 368-380. PICKETT, A.D. (1959) Utilisation of native parasites and predators. J. econ. Ent. la, 1103-1105. PONTIN, A.J. (1961) Prey of Lasius niger and Lasius flavius. Ent. mon. I1ag. 22, 135-137. POULTON, E.B. (1906) Predacious Insects and Their Prey, Pt. 1. Trans. Ent. Soc. Lond. 1906, 323-409. PO Y] J.L. & STAGE, G.I. (1962) Prey selection by robber- flies of the genus Stenopogon, with particular observa- tions on S, engelhardti Bromley (Diptera:Lsilidae). 176.

Wasmann. J. Biol. San Francisco 20, 139-157. PUTMAN, W.L. (1955) Bionomics of Stethorus punctillium Weise in Ontario. Canad. Ent. q.2, 9-33. PUTMAN, I.L. (1957) Laboratory studies on the food of some Coccinellids found in Ontario peach orchards. Canad. Ent. 12, 572-579. RABAUD, S. (1926) Sur la regime alimentaire des larves de Chrysopa vulgaris. Feuill. Nat. Paris A2, 164-167. ROSS, W.A . & ROBINSON, (1922) Notes on the Plum Spider Mite or European Red Mite. 52nd. Ann. Rept. Ent. Soc. Ontario. 1921, 33-42. SCHILDER, F. & SCHILDER, M. (1928) Die Nahrung der Coccinelliden and ihre Beziehung zur Verwandtschaft der Arten. Arb.Biol.Reich.Land-Forst. Berlin. 16, 213-281.

SCHRODER, C. (1904) Bericht Aber die wAhrend des Jahres 1903 zur Einsendung gebrachten Schadlinge. Landw. Tfochenbl. f.. Schleswig-Holstein. 32, 440-443. SHELFORD, V.E. (1909) Life Histories and Larval Habits of the Tiger Beetles. J. Linn. Soc. Lond. 22, 157-184. SNEEI C. (1922) British Ladybird Beetles. Their control of Aphids. Fruitgrower. 3, 675-6,717-8 and 759-760. SEIRNOFF, 7.A. (1958) An Artificial Diet for Rearing Coccinellid Beetles. Can. Entom. 20, 563-565. 177.

SMITH, B.C. (1960) A technique for rearing Coccinellid beetles on dry foods, and influence of various pollens on development of Coleomegilla maculata lengi Timb. Can. J. Zool. 28, 1047_1049. SMITH, B.C. (1965a) Growth and Development of Coccinellid larvae on dry foods. Canad. Ent.,,, 760-768. SMITH, B.C. (1965b) Effects of food on the longevity, fecundity and development of adult Coccinellids. Canad. Ent. 22, 910-919. SMITH, B.G. (1965c)Differences in Anatis mali Auct.and Cabomegilla maculata lengi Timb. to changes in the quantity and quality of larval food. Canad. Ent. In Press. SMITH, H.D. (1936) Le Syrphide Xanthandrus comtus Harris predateur de la Chenille Processionaire du Pin. Bull. ent. Soc. France. 41, 328-329. SMITH, K.G.V. (1952) On the prey of trio swarms of Hilara maura (Diptera:Empididae). Ent. mon. Mag. 88, 38-39. STROITHAL, H. (1927) Die Larven der palaearktischen Coccinellini and Psylloborini. Arch. Naturg. 42A, 1-63. S-IEETHAN, H.L. (1936) The Biological Control of Insects. Neu York, 461pp. S2UMX072SHI, 17. (1951) Observations on Coccinellidae II. 178.

Experimental rearing of Coleomegilla on a non-insect diet. Trans.Int.Cong.Ent.9th. Amsterdam. 1, 781-785. TAN, Chia Chen (1934) Notes on the biology of the lady beetle Ptychantis axiridis Pallas. Peking Nat.Hist. Bull. 8, 9-18. THOMPSON, U.R.(1928) On the relative value of parasites and predators in the biological control of insect pests. Bull. ent. Res. 12, 343-350. THOMPSON, U.R. (1951) The Specificity of Host Relations in Predacious Insects. Caned. Ent. L32, 262-269. UATSON, J.R. & THOMPSON, 7.L. (1933) Food Habits of Leis conformis Boisd. Fla. Ent. 22, 27-29. WHEELER, 7.M. (1913) A solitary wasp (Lphilanthrops frigidus Smith) that provisions its nest with queen ants. J. Lnim. Behaviour 1, 374-387.

1:TIMIDLER, U.M. (1926) Ants. Their Structure, Development and Behaviour. New York, 661pp. WINHLER, J.R. (1952) Coccinella 7-punctata as a predator on Ameletus inopinatus (Ephemeroptera). Acta Soc.ent Cechoslov. Praha. 2, 242-243. UITHYCOLIBE, C.L. (1923) Notes on the Biology of some British Ileuroptera. Trans. Ent. Soc. Lond. 1922, 501-594.