The Digestive Enzymes of Some Wood-Boring Beetle Larvae

The digestive powers of insects living in wood have long been a subject of interest, but it is only during the last ten to twelve years that a real attempt has been made to investigate the physiology of digestion of these organisms. The lack of research upon this problem is probably to be explained by the difficulty of obtaining adequate supplies of insects and of experimenting with organisms feeding upon a substance of such complex chemical constitution. A wide field for investigation still remains, but sufficient work has been published to enable the trend of the results to be appreciated. It is of the utmost importance that research should be undertaken upon as many species of wood-borers as possible, since it is already evident that there exists considerable diversity in the ways in which insects derive their sustenance from wood.

Uvarov (1928) in his summary of the literature up to 1927 on insect nutrition concluded that the food relationships of the vast majority of wood-eating insects were absolutely unknown, but he overlooked two important papers by Mer (1893, 1903), who had produced evidence to show that certain beetle larvae (Lyctus spp.) are dependent upon the starch in wood. Another review of knowledge of the digestion of wood by insects was made in 1934 by Mansour & Mansour-Bek (1934b), who were able to consider, in addition to their own experiments (1934 a), the results of several investigators who had worked in the interim. They distinguished between two types of wood-eating insects, those secreting a cellulase and those unable to do so. The former derive nourishment mainly from the cell-wall components, whilst the latter are dependent upon the cell reserve materials. More recent information (Hopf, 1938) suggests that wood-feeding beetle larvae may be divisible into three classes on the basis of the complexity of the wood components which they hydrolyse during digestion.

In spite of the work so far carried out, however, only the fringe of the problem of wood-feeding has been investigated. For instance, the restriction of many species of wood-borers to a limited range of hosts has not yet been explained and information upon the relation between larval nutritional requirements and host specificity must be of interest. In this connexion the demands of the larva for a supply of suitable food material may be reflected in the tropisms of the beetle when selecting a site for oviposition. For example, Parkin (1936) has demonstrated that the need of Lyctus larvae for starch is met by the ability of ovipositing females to recognize the presence of this substance in wood and that, in consequence, eggs are not laid in starch-free timber. Furthermore, the role of symbiotic microorganisms in digestion is not yet settled beyond question. Unlike the Protozoa in the digestive tracts of most Termites, the symbiotic yeasts in wood-boring beetle larvae appear to play no direct part in the digestion of wood, but Koch (1934) has made the very interesting discovery that larvae of Sitodrepa panicea L., an Anobiid living on stored foodstuffs, are greatly stunted in growth when artificially rendered free from symbionts, and that normal growth can be restored by addition to the food of yeast or wheat germ. If the active principle of the yeast and wheat should prove to be a vitamin, the knowledge may prove valuable in relation to experiments upon the vitamin requirements of wood-borers, a subject which has not yet been investigated.

Up to the present, research upon the food relations of wood-boring coleopterous larvae has followd four main lines, namely:

• (i) chemical analysis of food and excrement;

• (ii) removal of certain substances from the wood by extraction with solvents ;

• (iii) feeding on artificial diets ;

• (iv) testing for the presence of enzymes in the gut.

These experimental methods are clearly complementary and the proper correlation of the results obtained should yield the answer to the question, how can insects satisfy their nutritional requirements from a diet of wood. The value to applied entomology of such fundamental knowledge is illustrated by the work of the author and others upon the food relations of Lyctus powder-post beetles, wood-borers of considerable economic importance. It has been shown that starch in wood forms the principal food material for the larvae and that female beetles avoid oviposition in starch-free timber. This knowledge has led to research upon methods of rendering wood free from starch as a means of preventing infestation by Lyctus.

The investigation dealt with in this paper has been concerned with tests for the presence of certain enzymes in the guts of a number of species of wood-boring beetle larvae, and an attempt has been made to relate the results to our present knowledge of the chemistry of wood and to the work of other investigators. The experiments have been confined to Coleoptera able to feed in bark or wood which is sound or but slightly decayed, and the term “wood-borer”, for the purpose of this paper, does not include such forms as can live only in wet wood in an advanced stage of fungal decay. There is, of course, no hard and fast line of demarcation between the two groups.

Before dealing with the experimental side of the work, it seems advisable to give a brief account of the composition of the substrate upon which wood-boring insects feed. Chemical analysis of wood is usually undertaken according to a series of standard methods which enable the amounts of the principal components to be measured. The meaning of the resulting data is often not clearly understood by biologists owing to the difficulty of realizing the relation between substances isolated or estimated by chemical methods and their counterparts incorporated in the structure of the wood. For our purpose, we may consider wood to be formed of minor and major components according to whether they are present in small or large amounts.

The minor components include the cell contents, which are generally removable from finely divided wood by neutral solvents and consist of such materials as tannins, resins, dyes, alkaloids, fats, oils, gums, sugars, starch, proteins and mineral salts. By prolonged extraction with boiling water, Campbell (1935) has isolated from oak sapwood substances which he terms “amylo-uronides”. These are hydrolysed by takadiastase to yield glucose, but they also contain some xylose and uronic acid units and are probably intermediate in composition between starch and certain hemicelluloses.

The major components, which usually comprise about 90 % of the wood substance, are cellulose, lignin and hemicelluloses. The modem biochemical conception of the fine structure of the cell wall is explained in a review by Clarke (1938). Cellulose molecules are aggregated into micelles which may be regarded as the basic units forming the framework of the cell wall. On this framework are encrusted lignin and hemicellulose in such a way as to form a second continuous and interpenetrating system. These two systems cannot at present be separated by chemical analysis without causing some constitutional change in the desired product.

In the present state of knowledge, the cellulose molecule is to be regarded as a chain of anhydro-glucose units united through β-linkages and starch as the corresponding substance with α-linkages. Owing to its more symmetrical spatial configuration, the cellulose molecule is more stable than that of starch. In spite of the large volume of research which has been carried out to ascertain the constitution of lignin, little more is known than that it is probably largely aromatic in character. Similarly little is known about the composition of the hemicelluloses. Some appear to be closely associated with the cellulose and in most wood analyses are estimated as “pentosans in the cellulose”. O’Dwyer (1937, 1939, 1940) has recently been able to show that hemicelluloses A and B of oak heartwood have a common recurring structural unit consisting of anhydroxylose and methylhexuronic anhydride units in the proportion of 6:1 respectively. The corresponding products from sapwood have as an additional constituent a variable number of anhydroglucose units. There must presumably be some difference in degree of aggregation of the molecules of the two hemicelluloses to account for the greater solubility of B, which should render it more easily available to wood-feeding insects.

Chemical analysis unfortunately does not enable the absolute quantities of the various components as they occur in the wood to be estimated. For instance, the isolation of cellulose by delignification of wood with chlorine and sodium sulphite yields “Cross and Bevan cellulose”, a product containing α-, βand γ-celluloses. There is no evidence to show that β -and γ-celluloses occur in such as wood: in fact and γ-celluloses are probably derived from α-cellulose during the chemical treatment. Similarly, hydrolysis of all the carbohydrates of wood with 72 % sulphuric acid leaves a residue called lignin by wood chemists, but which is generally held to be different from the lignin occurring naturally in the cell wall. However, by the use of standardized methods, chemical analysis gives results which are capable of repetition and can therefore be used to demonstrate changes in the composition of wood and to indicate which components have been most affected.

One of the main methods of approach to the problem of wood feeding has been the comparison of chemical analyses of the wood in which an insect is tunnelling and the bore-dust remaining in the galleries. This method reveals whether there has been any change in the proportion of the various wood components during the process of digestion, but cannot show the amount of each component eaten unless the weight of the wood sample is known before and after attack. Most workers have tried to overcome this difficulty by assuming that the ash or lignin content of the wood is not affected during passage through the larval gut. From spectrographic analysis and X-ray photography of Lyctus larvae in wood, however, Jones & Ritchie (1937) have shown that mineral constituents are absorbed by the insects. Furthermore, it is well known that wood-destroying fungi can break down lignin, and until proof is available that insects cannot also do so, the assumption that the lignin is not altered during larval digestion is hardly justifiable. The chemical analyses so far published are not rendered valueless by being based on this assumption, however, since the figures obtained represent the minimum changes which have occurred. Proof of the decomposition of lignin during digestion would therefore mean that a higher proportion of the other constituents is utilized than has hitherto been revealed.

The larvae used in this investigation were obtained principally from the stocks of infested wood in the insectaries at the Forest Products Research Laboratory or from samples submitted for examination as part of the advisory work of the Laboratory. Identification of the insects was based upon knowledge of the type of injury caused in timber, examination of the morphological characters of the larvae and comparison with such descriptions and keys as are available, and, where possible, identification of beetles associated with or reared from the larvae. Data on the species used and the origin of the timber in which they were found are summarized below.

Throughout the investigation only qualitative tests for digestive enzymes were undertaken and the technique finally adopted was based on work by Swingle (1925) and Wigglesworth (1927). Special attention was paid to the occurrence of carbohydrases; a test for proteinase was also carried out, but no satisfactory test for ligninase applicable to small quantities of insect digestive juices could be discovered.

Various methods have been used by different workers in order to detect the presence of a cellulase in the alimentary canal of wood-boring larvae, and some of the results must be accepted with caution, pending confirmation. The reliability of the test involving breakdown of filter-paper or cotton-wool is dependent upon the purity of the material with regard to a-cellulose content. Moreover, unless the difference is considerable, it is not safe to assume the presence of a cellulase from cellulose determinations of food and frass by the Cross and Bevan method of analysis, owing to the presence in the cellulose of pentosans which may occur as impurities in quantities up to 20 %, especially in hardwoods. The occurrence of so high a proportion of pentosans in Cross and Bevan cellulose also precludes the use of this substance as a substrate for tests with enzyme solutions. The most reliable tests would appear to be the breakdown of a-cellulose in vitro or the disappearance of cellulose from sections of plant tissue as shown by the chlor-zinc-iodide reaction; a-cellulose prepared from oak was therefore used as the substrate in the present work.

Larvae were cut from the wood not more than one day before they were required, 10-150 being used for each experiment according to size. The guts were dissected out and freed, so far as possible, from fat body, tracheae, Malpighian tubes, etc. They were then transferred to an agate mortar containing about 3 c.c. of previously boiled distilled water to which a drop of toluene had been added. The guts were broken up by teasing with needles and were then ground with a small quantity of clean sand. The sand was allowed to settle and the supernatant fluid pipetted off: more water was added and the process repeated until about 10 c.c. of dilute digestive fluid and tissue suspension were obtained. The liquid was divided into two equal portions, one of which, after heating in a boiling water-bath for 15 min. to inactivate the enzymes present, was used as control. As soon as they were prepared, both suspensions were covered with a film of toluene.

Glass tubing, 3 mm. in internal diameter, was drawn out as shown in Fig. 1 to form at one end an ampoule about 2-5 cm. long. By means of a rubber teat, the ampoule was one-third filled with enzyme suspension, when substrate (1 % soluble starch, 3 % sucrose, 3 % maltose, or 3 % lactose) was sucked in until only a small air bubble remained. After breaking off the extension of the ampoule bearing the rubber teat, the ends were sealed in a flame, and by repeated inversion and shaking the bubble was made to mix the contents thoroughly. With solid substrates (a-cellulose, hemicelluloses A and B^*) a small quantity was inserted into the ampoule after drawing out the first end. A similar set of ampoules was made with the boiled fluid to act as controls. Both sets were then maintained at 37° C.

The preparation of 10 c.c. of suspension allowed each test to be made in duplicate, one of the ampoules being opened after approximately 70 hr., and the other after 140 hr. After breaking off the ends of the ampoule, the contents were expelled into a small test tube and tested for the presence of reducing sugars by boiling with Fehling’s or Barfoed’s reagents. The intensity of copper reduction, hence the activity of the ferment, was recorded approximately by one to four crosses (see Table I). It must be remembered that comparison of the activity of the enzymes in the different species of insects is not strictly permissible owing to variation in the numbers of larvae used in each experiment.

The detection of proteinases was undertaken by the photographic gelatin method described by Parkin (1936).

Considerable dilution of the enzymes occurs during extraction as described above, and the clear-cut results of their action which were generally obtained during the investigation may be taken as evidence of the great activity of these substances at their normal concentration in the gut. The poor hydrolysis of a-cellulose in the tests may be due to the probable increase in resistance to hydrolysis of cellulose when submitted to the action of strong reagents during isolation from wood. It should also be noted that the hemicellulose A used in the tests was extracted from oak sapwood and contained some 10 % of glucose residues, whereas the hemicellulose B was prepared from heart-wood and was free from hexosan.

Table I shows the results of the experiments to detect certain enzymes in the digestive tracts of a number of species of wood-boring beetle larvae. When sufficient insects were available, results were confirmed by repetition of the tests. The activity of the enzymes sometimes varied slightly in different batches of larvae of the same species and the value included in the table represents the highest recorded.

For the sake of convenience the results of this investigation together with those of other investigators will be considered family by family.

Beetles of the family Scolytidae may be classified in two groups of entirely different habits according to whether the life cycle is passed in the solid wood (Ambrosia beetles) or in the bark (bark beetles). Members of the first group are dependent for nourishment upon fungi growing on the walls of their galleries but nothing is known of the physiology of their digestion. In the second group the only species upon which information is available is the ash bark beetle, Hylesinus fraxini Panz. Hopf (1938) has shown by tests for digestive enzymes that the larvae probably feed upon soluble sugars, starch, the hexosan part of the hemicelluloses, and proteins. These results he confirmed, so far as possible, by chemical analysis of the food and frass, assuming, in the absence of a positive result in his tests for a cellulase, that cellulose is not hydrolysed. He reports the absence of an enzyme capable of hydrolysing the hemicellulose A of ash bark and claims that, so far as hemicelluloses are concerned, the larvae are unable to break down the pentosan and can utilize only the hexosan part of hemicellulose B.

The detection during the present work of amylase, invertase, maltase, lactase, and proteinase in Phloeosinus bicolor, together with a negative result for cellulase, agrees with the findings of Hopf. However, an enzyme which strongly hydrolysed hemicellulose A prepared from oak sapwood was also found, and since, according to O’Dwyer (1937), this hemicellulose contains only 9-10 % of anhydroglucose residues, a considerable portion of the pentosan fraction must also have been hydrolysed. The utilization of pentosans is confirmed by the very strong action of the digestive juice upon hemicellulose B which was prepared from oak heartwood and contained no starch or other hexosan.

No mention is made in the table of Scolytus destructor, since the result of tests with a tissue suspension of the larval guts was entirely negative. This is extremely interesting, as the larvae were fully grown and hibernating: many of the larvae pupated after being kept in a warm room for three days. Wood was not present in the larval guts nor was there any appreciable quantity of digestive juice, and it may therefore be concluded that, at the start of the diapause with consequent cessation of feeding, the larvae had ceased to secrete digestive enzymes.

The large size of the larvae has probably induced several workers to concentrate on this family. Seillière (1905) detected an enzyme hydrolysing xylan in the alimentary canal of Phymatodes variabilis L. Falck (1930a) and Horn (1930) have analysed the food and frass of Hylotrupes bajulus L. and shown that during digestion depletion of cellulose and pentosans occurs. More recently, Schuch (1937) has shown that some constituent of the cell contents is necessary for the best growth of the larvae of this insect, since they increase in weight much more rapidly when feeding in the outer zone of the sapwood than in the inner zone or in heartwood. In 1938, Becker observed a considerable acceleration of growth when larvae were inserted into wood impregnated with peptone solution, but not when into wood impregnated with soluble starch or glucose: an intermediate value was obtained with an extract of malt. He suggests that the amount of protein in wood may be a governing factor in the growth of the larvae and that a small amount of nitrogenous matter was present in the malt extract. Far more striking, however, are the results reported by Gösswald (1939) who found that, 180 days after hatching, larvae in wood impregnated with 5 % diastase solution weighed up to 114 mg., while similar larvae in peptone-wood weighed up to 4 mg. and in the untreated controls up to 2 mg. No explanation is offered for this phenomenon. It seems that it cannot depend upon the enzymic action of diastase on the wood starch, since the quantity present in softwood timbers is very small, but it is possible that the diastase contains as an impurity some growth-promoting substance the presence of which enables the larva to utilize larger amounts of some or all of the major components of the wood.

Ripper (1930) has demonstrated the presence of a cellulase in the mid-gut fluid of larvae of Cerambyx cerdo L., Rhagium bifas datum F., and Leptura sp. In 1934 Müller showed by chemical analysis that there is a considerable loss in cellulose and pentosans from wood passing through the larval gut of Oxymirus cursor L., Leptura rubra L. and Gracilia minuta F. and also detected in the alimentary canal of Oxymirus cursor the enzymes cellulase, hemicellulase, xylanase, amylase, invertase, maltase, emulsin, lipase, trypsin and erepsin. Mansour & Mansour-Bek (1934a, 1937) have shown also that the larvae of Macrotoma palmata F. and Stromatium fulvum Vill. possess an enzyme hydrolysing cellulose, but in the first paper record the discovery of a Cerambycid larva, Xystrocera globosa 01., which cannot secrete a cellulase and appears to derive its food from the minor carbohydrate constituents which occur in plenty in the sapwood in which it lives. Finally, Pochon (1939) has isolated a cellulolytic bacterium from the larval digestive tract of Rhagium sycophanta Sch. living in rotten wood and states that it is probably responsible for the digestion of cellulose in this species, but this assumption is open to the criticism that the bacteria may perhaps occur normally in decaying wood and are merely incidental in the gut of the insect.

As shown in Table I, the results of tests upon six species of Cerambycid larvae proved very uniform. The only comparable investigation, made by Müller (1934) upon the larvae of Oxymirus cursor, gave results in close accord with those of the author. In general, therefore, it may be concluded that the majority of Cerambycid larvae digest the starch, soluble sugars, hemicelluloses, cellulose and proteins in wood.

It is noteworthy that there is no fundamental difference between the digestive powers of species feeding upon bark, phloem and cambium, e.g. Rhagium, Phymatodes, and those living in the solid wood, e.g. Isotomus, Xylotrechus, Hylotrupes, Smodicum, although, according to the one test undertaken, larvae of the last named insect seem to be deficient in enzymes hydrolysing the cell content carbohydrates.

The number of genera of this family in which a cellulase has been reported now amounts to twelve, the only exception so far recorded being Xystrocera globosa which, according to Mansour & Mansour-Bek (1934a), lacks a cellulase and is dependent for its food upon cellcontents.

Mer as early as 1893 intimated that starch in timber was the principal food of Lyctus larvae. In 1929, Campbell reported that chemical analyses of food and frass showed no difference in the proportions of the major components and concluded that the larvae must derive their nourishment solely from the cell contents. Wilson (1933) has since confirmed Mer’s statement that Lyctus cannot develop in starch-free wood. Parkin (1936) concluded that the larvae require starch, protein and an unidentified water-soluble substance, and also succeeded in rearing young larvae to beetles upon an artificial diet of starch, sucrose and peptone in the complete absence of wood. In addition, he detected in the larval gut the ferments amylase, invertase, maltase, lactase and proteinase. The presence of these enzymes and the absence of any capable of hydrolysing cellulose or hemicellulose A have been confirmed during the present series of tests. A new result, however, is that the digestive fluid of the larvae can cause a partial hydrolysis of hemicellulose B, but whether this can be done when the substrate is incorporated in the structure of the cell wall requires further investigation.

The ease of detection of starch in wood by means of a dilute aqueous iodine solution makes it possible to obtain some additional information on the action of the amylase. Microscopical examination shows the wood particles in the mid-gut of Lyctus larvae to be so small that nearly every cell is broken open, yet starch grains can often be found in the rectal contents and in the bore-dust in the tunnels. There is always less starch in the frass than in the original wood but the percentage reduction is very variable, suggesting that the activity of the amylase differs in different larvae. In abundantly starchy wood, Lyctus bore-dust sometimes contains a considerable quantity of starch and, when tunnels cross, less starch is often detectable at the point of intersection, indicating that further nourishment is obtainable from the starchy frass when eaten by a second larva.

Little is known of the food requirements of this family, closely related to the Lyctidae, but Beeson & Bhatia (1937) in their recent summary of knowledge on the biology of the Indian species state that starch is an essential constituent of the food of Dinoderus, Heterobostrychus and Sinoxylon, if not of all Bostrychidae. With reference to attack by Dinoderus spp. in bamboo, they summarize the results so far obtained as follows :

“An attempt to reconcile the data and explain the contradictions and inconsistencies of the experimental work… is foredoomed to failure ; but one is left with the impression that (a) starch is an essential food-substance, (b) glucose is not essential, (c) a soluble substance possibly a disaccharide is an essential, and that (d) neither of the essentials alone is sufficient.”

The results of the present investigation show that the larvae of Heterobostrychus and Bostrychoplites feed exclusively upon the cell contents, namely starch, sugars and proteins. Unlike Lyctus larvae they are apparently unable to utilize any portion of hemicellulose B.

According to chemical analyses of wood and frass by Falck (1930 b) and Müller (1934), Anobium punctatum larvae can utilize cellulose and hemicellulose. In the case of Xestobium rufovillosum, Campbell (1929) reported, after analysing sound oak heartwood and larval frass, that the larvae of this beetle bring about a change in the carbohydrate : lignin ratio of oak which indicates utilization of the cell wall carbohydrates. Similar analyses have been performed by Ripper (1930) and Norman (1936). None of these authors has taken into consideration in the interpretation of his results the change in composition of wood brought about by fungal decay, which has been shown by Fisher (1935) to be a necessary preliminary to Xestobium attack. Ripper, however, confirmed his conclusion that cellulose was digested by detecting the presence of a cellulase in the larval gut fluid.

Four members of this family have been investigated and a positive result obtained for each enzyme test undertaken. Therefore the larvae of Anobium, Xestobium, Ernobius and Ptilinus are able to break down the carbohydrates of the cell wall, namely cellulose and hemicelluloses, as well as to utilize the protein and, when they are present, starch and sugars in the cell contents.

On the basis of their nutrition wood-eating insects have been divided by Mansour & Mansour-Bek (1934a) into two types, namely those which have no cellulase and can consequently live only in timber comparatively rich in starch and sugars, and those able to secrete a cellulase and therefore to live in timber relatively poor in these cell contents. The results of the present work indicate, however, that there are three groups into which the wood-boring Coleóptera may be divided on the basis of their digestive powers as follows:

• (i) Larvae able to utilize only the cell-contents and perhaps part of the polysaccharides which are intermediate in composition between starch and the hemicelluloses—Lyctidae and Bostrychidae.

• (ii) Larvae able to utilize cell contents and the carbohydrates of the cell wall up to hemicelluloses, but excluding cellulose—Scolytidae (bark beetles).

• (iii) Larvae able to utilize cell contents and the carbohydrates of the cell wall including cellulose—Anobiidae and most Cerambycidae.

The position of the Cerambycid, Xystrocera globosa, which is reported by Mansour & Mansour-Bek (1934a) to lack a cellulase, is not clear as these workers did not determine whether the larva could digest hemicelluloses.

It is interesting, for the purposes of comparison, to note that at least two types of fungal attack occur in wood. Most of the surface moulds appear to live solely upon cell contents and their hyphae pass from cell to cell through natural openings such as pores. In the wood-rotting Polyporaceae, twenty-six species of which have

The Digestive Enzymes of some Wood-boring Beetle Larvae 375 been investigated by Bose & Sardar (1937) and Garren (1938), ferments causing the breakdown of cellulose, hemicellulose and lignin are of fairly general occurrence. These Basidiomycetes extract nourishment from the cell wall and pass from cell to cell by perforating the wall itself through enzymic action.

There can now be no doubt of the existence of a cellulase among the digestive enzymes of many wood-boring beetle larvae, although it is not of universal occurrence. The view advanced by several authors that the digestion of cellulose by coleopterous larvae is dependent upon a symbiosis with micro-organisms which are responsible for secretion of cellulase has now been shown by Müller (1934) and Mansour & Mansour-Bek (1934a) to be untrue, at least with reference to the Cerambycidae and Anobiidae. However, in the light of Koch’s (1934) work on Sitodrepa panicea and similar research on other insects which are not wood-borers, it is possible that the organisms are important in connexion with the vitamin requirements of the insects.

According to Fraenkel (1936) and others, the extent of utilization of pentose sugars varies considerably among insects but, in view of the strong action of their digestive juices upon the hexosan-free hemicellulose B of oak heartwood, it must be accepted that the larvae of many wood-boring Coleoptera can use them in their metabolism.

The source of nitrogen for insects feeding on wood has long been a subject of discussion, since some workers have thought that wood was lacking in protein. Heitz (1927) suggested that the internal symbionts might be able to fix atmospheric nitrogen, but Müller (1934) has been unable to find any evidence to support this as a result of artificial culture of the organism. Ripper (1930) and Mansour & Mansour-Bek (1934b) assume that insects can derive their nitrogen supply directly from the wood, but no attempt had been made to determine whether proteolytic enzymes occur in the alimentary canals of wood-boring insects until Parkin reported in 1936 the discovery of a strong proteinase in the digestive juice of Lyctus larvae. At the same time he pointed out that in wood attacked by Lyctus the carbohydrate : protein ratio of the cell contents is such that the larvae might be expected to obtain their supply of nitrogen from the wood without difficulty. Hopf (1937) has since shown that the larvae of Hylesinus fraxini, Lyctus sp., and Anobium punctatum cause a reduction in protein nitrogen of the wood as it passes through their intestines and has detected (1938) trypsin and erepsin in larvae of Hylesinus. The present demonstration of a proteinase in the larval guts of fourteen species of wood-boring Coleóptera belonging to five families indicates that such a ferment occurs generally and that the larvae are not dependent upon symbiotic microorganisms for their nitrogen supply.

According to the results of this investigation there would seem to be no reason why some of the species examined should not be capable of deriving nourishment from wood known to be unsuitable. For instance, Hylotrupes bajulus confines its attack to coniferous timbers and is unknown in hardwoods although the enzymes present in the larval gut would apparently fit it to digest either kind of wood with equal facility. Similarly, many borers are recorded only in hardwoods and some insects are confined to one or very few species of timber. Furthermore, no explanation has yet been found for the restriction of certain insects, e.g. Anobium punctaium, to seasoned wood when it might be supposed that they could feed equally well in unseasoned wood and vice versa. It is clear therefore that the nutrition of wood-boring Coleóptera depends upon more factors than the enzyme complex in the larval gut and much research remains to be carried out before any comprehensive idea can be gained of the interaction of the various conditions governing the food relationships of wood-eating insects.

The work described above was carried out as part of the programme of the Forest Products Research Board and is published by permission of the Department of Scientific and Industrial Research. The author wishes to thank Dr R. C. Fisher, officer in charge of the Entomology Section, and Mr W. G. Campbell, officer in charge of the Wood Chemistry Section, for their helpful criticism of the manuscript.