Chemical ecology of bark beetles (Scolytidae)

Byers, J.A. Chemical Ecology of Bark Beetles. Experientia 45:271-283. pdf

CHEMICAL ECOLOGY OF BARK BEETLES

John A. Byers

Dept. of Ecology, Animal Ecology, University of Lund, S-223 62 Lund, Sweden.
Present address:

Summary: The purview of chemical ecology and the recent criticisms of improper application of theory to bark beetle phenomena is briefly discussed. Seven levels of research in chemical ecology are presented as well as their relationship to research on bark beetles. The biology and chemical ecology of several pest bark beetles from North America and Europe are discussed in regard to host-tree selection theories of random landing on trees or attraction to semiochemicals. The diversity and similarities of pheromone components among species are presented in relation to their biosynthesis from host-tree precursors and in relation to the ecological implications of de novo or precursor syntheses. Individual variation in biosynthesis of, response to, and release of pheromones is discussed. Olfactory perception of semiochemicals at both the electrophysiological and behavioral levels is presented. Orientation to semiochemicals during walking and flying is discussed with reference to the significance of dose-response curves for determining a compound's functionality in short- or long-range communication. The regulation of attack density, termination of the aggregation, mechanisms of attack spacing, and recognition of host suitability are presented in the context of an individual's avoidance of intra- and interspecifc competition. Finally, a brief summary of topics where our understanding of the chemical ecology of bark beetles and their associates is poorly known is presented.

Key words. Scolytidae, semiochemical, pheromone, allomone, kairomone.

Chemical ecology and ecology

Ecology is the science of the relations of an organism to both its biotic and abiotic environment which influence the organisms' distribution and abundance⁷⁷. The biotic factors are included in the disciplines of physiology, behavior, genetics and evolution; ecology is especially concerned with the interface of these areas. Chemical ecology then concerns any aspect of ecology but involves the external chemicals which mediate the interactions. This definition is actually more comprehensive than many would accept, for instance, it could include macrophage antibody - microbe interactions, nutrient cycling, and much of biology. Traditionally, chemical ecology has been restricted to studies of the chemicals (semiochemicals) which mediate interactions between individuals of a species (pheromones) or between co-evolved species (allelochemicals, such as kairomones and allomones).

Alcock¹ criticized several bark beetle (Scolytidae) researchers for what he thought was their inadequate application of ecological theory in the explanation of bark beetle phenomena. His primary concern was that bark beetle mass-attack and colonization of host trees was often misunderstood in terms of "species- selection" when in fact "individual-selection" was now the dominant theory. Today most realize that a deeper knowledge of bark beetle biology can be obtained if both the proximate and ultimate causes for the phenomena are considered.

Levels of research in chemical ecology

Research in chemical ecology of insects can be described with seven hierarchial levels which to some extent are chronological. The following examples will refer to the western pine beetle, Dendroctonus brevicomis, which aggregates en masse in response to an aggregation pheromone. The first step in the research is to (1) observe the biological phenomenon - e.g. demonstrate that D. brevicomis are attracted to odor from the infested host tree⁸⁹. Once the biology is partially understood, one can design bioassays which are used to (2) isolate and identify at least one semiochemical component - e.g. exo-brevicomin was isolated from female frass by solvent extraction, concentration, gas-liquid chromatographic (GC) fractionation (collecting into discrete fractions the continuous successive elution of chemicals from the GC) and bioassay (testing each of the several fractions for attractive activity)¹²⁹. Once a relevant compound is isolated, enough must be obtained for structure elucidation via GC-mass spectrometry (GC- MS), and often other spectrographic methods¹²⁷^-
1²⁹. The third step involves (3) isolating and identifying all participating semiochemical components. For example, D. brevicomis males were shown to produce frontalin⁷³ which was synergistic with exo-brevicomin (i.e. neither compound alone is very attractive but together the blend is highly attractive), and the host-tree monoterpene, myrcene, further enhanced the attraction¹¹^,¹².

Once one chemical part of the ecological system, in this case the aggregation pheromone, is described, then (4) other pheromones, which may also act as kairomones or allomones, can be identified with respect to the biological phenomena. For example, intraspecific pheromone inhibitors of attraction such as verbenone¹³^,¹⁰⁹ and trans-verbenol¹³^,³² cause individuals to avoid colonizing in high attack density patches and thus function in terminating the aggregation⁴⁵. To further understand the chemical ecology, which appears to grow ever more complex, one must (5) quantify the rates of production and release of semiochemicals over the period of colonization. Some studies have measured the release rates of some semiochemicals from D. brevicomis over short periods²⁷, and others have followed the production of pheromone components in the beetle's guts during the colonization period⁴⁵.

Once the release rates of semiochemicals are known, the (6) mechanisms of pheromone production, olfactory perception, and the sites of biosynthesis and perception can be investigated - these areas will be discussed later. However, these topics become increasingly more peripheral to chemical ecology as they become more involved in physiology, behavior, genetics and evolution. Finally, one can (7) attempt a synthesis of the knowledge concerning known semiochemicals, release rates, interspecific interactions and biological phenomena in order to construct a working theory on how the semiochemicals mediate the interactions between individuals of various sexes and species.

Chemical ecology of bark beetles

Reviews of various aspects of bark beetle chemical ecology since 1980 have concerned semiochemicals in host selection and colonization¹⁴⁷, host odorants and pheromones⁷⁸, orientation²³, aggregation¹⁸, aggregation pheromones²²^,²³, stridulation and pheromones¹¹⁷, biosynthesis⁵⁹, and competition and semiochemicals³⁸. These reviews cover the field in more detail than can be done here. Instead this paper will highlight the chemical ecology of a few economically important systems: the pine beetles of the western and southern United States and the pine and spruce beetles of Europe in regard to selected topics.

The diversity of bark beetles and their biology can be appreciated in the taxonomic compilation of Wood¹⁵² and in discussion of mating systems⁷⁴. However, pest bark beetles are included in three categories: (1) "aggressive" phloem-feeding beetles which must attack en masse living trees and kill them in order to reproduce, (2) vectors of tree diseases such as Dutch elm disease, and (3) wood boring beetles which weaken and stain lumber. Beetles in all three groups utilize pheromones but much of the following discussion will best apply to those in the first group. A generalized life cycle of these bark beetles, including many non-pest species, is depicted in Fig. 1.

Fig. 1. Generalized life cycle of a pest bark beetle. The letters refer to the order of discussion in the text.

Before bark beetles begin their host-tree and mate seeking flight, there may be a period of required dispersal flight before they become responsive to pheromone⁴^,⁵¹^,⁶⁵. However, several species appear responsive to pheromone or host attractants soon after emerging from brood logs⁴⁰^,⁴⁶or overwintering sites⁴⁷.

Host selection. Little is known about dispersal and host selection except for the stages immediately preceding landing on the tree. Host selection for several pest species of the western United States is thought to be a random process in regard to landing on host and non-host trees⁹²^,¹⁵³. Ponderosa pines that were killed by freezing and screened to prohibit beetle attack, did not exhibit higher landing rates for Ips paraconfusus, D. brevicomis or D. ponderosae, among others, than did healthy trees. Landing rate differences also were not observed between healthy trees and trees with diseased roots. Other studies have shown that population densities of D. ponderosae⁷¹ and I. typographus ³^,⁴² are often high enough to allow nearly every tree in a stand to be visited by at least one beetle. If beetles test the defenses of host trees (although there is as yet no evidence of this), then, in theory, stronger trees will repel beetles while weaker trees will allow pheromone production and mass colonization. It is well known that conifers can be induced to exude copious amounts of resin which can entrap beetles⁴⁵^,⁸⁹^,⁸⁶^,¹²⁵ and has some toxic properties as well¹³¹. Healthy trees are able to exude more resin at a higher pressure¹⁴¹ while weaker trees, diseased or drought-stressed, are less able to produce resin. In late summer when trees in California are drought-stressed, it is often observed that attacks of D. brevicomis are not defended by resin flow and trees easily succumb⁸⁹. Thus, in many bark beetle species host selection appears to occur after landing on the bark, and in I. paraconfusus it was shown that a rejection of the non-host white fir did not occur until after penetration of the phloem⁵⁷.

In other species, however, there is a long-range influence of host-tree volatiles on host selection. D. rufipennis are attracted to volatiles from white spruce⁹¹, Scolytus multistriatus to volatiles from American elm¹⁰³, and D. valens to volatiles from ponderosa pine¹³⁹. Other bark beetles, often termed "secondary" because they infest trees in more advanced stages of decay or are sapwood-infesting species, have been shown to respond to volatiles from their host trees⁴⁹^,⁶⁴^,⁹⁰^,⁹² and especially to ethanol⁷⁶^,⁹⁰. Ethanol is the most commonly occurring kairomone attractant for the "secondary" species⁷⁶ and is supposed to be released from natural sources due to microbial activities, although this has not been quantified. In many cases it is known that a pheromone is released by the first arriving beetles which accounts for most of the subsequent aggregation of the population.

Also in contrast to a random landing process during host selection, Tomicus piniperda is strongly attracted by monoterpenes from Scots pine (terpinolene, alpha-pinene, and 3-carene) volatilizing from wound oleoresin⁴⁷. A GC fractionation and subtractive-combination bioassay was performed on volatile air collections of logs infested by T. piniperda in order to find an aggregation pheromone but none was detected, however, the host-tree monoterpenes (kairomones) were attractive⁴⁷. This phenomenon was also suggested to function in the beetle's recognition of host susceptibility. Fallen trees or broken tops injured during winter storms have wounds with oleoresin that release the attractive monoterpenes and these trees are less resistant to beetles since they have a reduced ability to produce more resin (lower turgor pressure). Ironically then, the beetle is selecting a susceptible tree based on the very components normally used by the tree as part of its resistance mechanism.

Once the first individuals of the "pioneer" sex arrive at the tree (males of Ips and Pityogenes or females of Dendroctonus and Tomicus) they are induced to begin boring into the tree. Little is known about what chemical stimuli are involved, few feeding stimulants or deterrents have been reported with the exception of S. multistriatus ⁵⁶^,⁶³^,⁹ ⁷. During the development of an artificial diet for I. paraconfusus it was found that sucrose stimulates feeding in a cellulose-based diet but not as much as sucrose plus 22% host-phloem particles in the diet⁴¹. Other studies have demonstrated feeding stimulant properties in extracts of host phloem ⁵⁸^,⁷¹. The pheromone is released from fecal pellets after the bark beetles have fed on the phloem (and some xylem tissue)¹⁴⁹.

Biosynthesis and release of pheromone. Ips species generally must feed for several hours in order to produce detectable attractants in their fecal pellets¹⁴⁹, and significant attraction may not result until males have fed for more than 24 hrs¹⁴⁹. The site of pheromone biosynthesis within bark beetles is still uncertain but the largest amounts of semiochemicals are found in the hindgut¹⁷^,²⁸^-
3²^,³⁹^,⁶⁹^,⁸⁰^,¹⁰⁴^,¹⁰⁷^,¹⁴ ². Male Ips acquire alpha-pinene from vapor absorbed during breathing in galleries and from ingested phloem, and then convert it to cis- verbenol (Fig. 2) or trans-verbenol depending on whether the precursor is of S or R chirality⁷⁵^,¹¹⁰.

Fig. 2. Biosynthesis of the tree monoterpenes (-)-alpha-pinene to corresponding enantiomers of cis-verbenol (an attractant pheromone component in Ips)^110,150, trans-verbenol (an inhibitory or attractant pheromone component in Dendroctonus)³² and possibly to verbenone (an inhibitory pheromone, allomone, and kairomone in several genera)^13,39; and conversion of the tree monoterpene myrcene to ipsdienol (pheromone component in Ips and Dendroctonus³⁰, ipsenol (pheromone component of Ips)^69,150 and probably (unproven) to amitinol (a pheromone component of I. amitinus)⁶⁰.

Dendroctonus species generally convert the respective S or R enantiomers of alpha-pinene to the corresponding enantiomers of trans- verbenol³². In contrast to trans-verbenol, the quantities of verbenone (Fig. 2) in male D. brevicomis are not affected by exposure to alpha-pinene vapors³².

Myrcene is converted to pheromone components hydroxylated at carbon 4 (ipsdienol and ipsenol) in various Ips species²⁹^,⁶⁹ or to amitinol in I. amitinus (Fig. 2)⁶⁰. The bio-conversion of myrcene to ipsdienol and ipsenol in I. paraconfusus is inhibited by the antibiotic streptomycin while other myrcene metabolites are not quantitatively affected.⁴¹ However, no direct evidence has been found for microbe involvement in pheromone production. The biosynthetic enzymes appear to vary between individuals, populations, and species since the enantiomers of ipsenol and ipsdienol vary quantitatively and qualitatively between individuals of a population⁷⁰^,¹³⁰ and between populations⁷⁹ of I. pini compared to other even larger differences between species¹⁴². D. brevicomis³⁰ and D. ponderosae⁷⁰ also are capable of transforming myrcene to (+)-ipsdienol which inhibits attraction to aggregation pheromone in D. brevicomis³⁰. In addition to the commonly used alpha- pinene and myrcene derived pheromone components, bark beetles in the genera Dendroctonus, Scolytus and Trypodendron utilize bicyclic ketals and tricyclic acetals (Fig. 3)⁷³^,⁸⁷^,¹ ⁰⁴^,¹²⁹.

Fig. 3. Structural similarities of Dendroctonus pheromone component enantiomers (frontalin, exo-brevicomin and endo- brevicomin)²⁹ to those of Scolytus multistriatus (multistriatin)¹⁰⁴ and Trypodendron lineatum (lineatin)⁸⁷.

In Fig. 3 it is easy to see the structural similarities of certain enantiomers of frontalin, exo-brevicomin and multistriatin. The importance of the correct enantiomer for each component in a pheromone was clearly demonstrated by Wood et al.¹⁵¹ wherein only (-)-frontalin and (+)-exo- brevicomin were active in causing D. brevicomis to aggregate. The biosynthesis of these components is uncertain but appears different between the sexes; in D. brevicomis the female produces exo-brevicomin after feeding whereas the male contains the largest amounts of frontalin upon landing on the tree⁴⁵.

Other structures of pheromone components of bark beetles are shown in Fig. 4.

Fig. 4. Examples of other pheromone components in Ips typographus (2- methyl-3-buten-2-ol)⁸, Ips cembrae (3-methyl-3-buten-1- ol)¹¹¹, Scolytus multistriatus (4-methyl-3-heptanol)¹⁰⁴, Gnathotrichus sulcatus (S-(+)-sulcatol)^24,48, Dendroctonus rufipennis⁶² and D. pseudotsugae, (R-seudenol)^83,145 and (3,2-MCH)¹¹⁵, and Pityogenes chalcographus (S,R- chalcogran⁶¹ and (E,Z)-(2,4)-methyl decadienoate^44,46).

Short chain terpenoid compounds are found in some European species (I. typographus, I. cembrae)⁸^,¹¹¹ but not in the most important American pest species. 3-Methyl-2-cyclohexen-1-one (MCH) is a well known inhibitor of D. pseudotsugae aggregation (Fig. 4)¹¹²^,¹¹⁵. A spiroketal, chalcogran, is found in several species of Pityogenes⁶¹^,¹²³, while recently a unique acetogenic pheromone component, E,Z-2,4-methyl decadienoate (E,Z-MD), was found in P. chalcographus (Fig. 4)⁴⁴^,⁴⁶. The biosynthesis of beetle pheromones, with many examples for Scolytidae, has recently been reviewed by Vanderwel and Oehlschlager¹³⁸ in regard to biosynthetic pathways (largely speculative) based on a suspected precursor. Much of the speculation concerns the origin of the pheromone components - whether they are terpenoid (Fig. 2), fatty acid, polyketide (Fig. 3), hydrocarbon, or amino acid (2- phenylethanol). The physiological mechanisms of biosynthesis and hormonal control are also reviewed. However, the ecological implications of the use of host-plant precursors compared to the de novo synthesis of pheromone components is of interest here.

The advantages to an individual using de novo synthesis of a pheromone component is that control of the stereo configuration and quantity can be precisely regulated resulting in the optimal benefit in the specific environmental context. The possible cost is that more biosynthetic machinery is necessary with additional costs of energy required for linking the simple building blocks. The advantages in using an exogenous precursor in which only a small change, such as hydroxylation, is needed to make a pheromone are obvious from an energetic standpoint. Also it seems that alpha-pinene and myrcene are routinely de-toxified by increasing their water solubility (hydroxylation or oxidation, Fig. 2), so the products were available in evolutionary time to be used as pheromones. The risk for the beetle is that the host-tree could influence the availability of monoterpene precursor or that variation in these compounds would confer a degree of resistance to those hosts with lower titres of precursor³⁰. Sturgeon¹³⁵ and Smith¹³²^,¹³³ have shown how host precursors of pheromones vary geographically in ponderosa pine and the question is whether tree resistance might also vary due to pheromone quantity or, as they suggest, due to toxic resistance properties of monoterpenes, especially limonene¹³³^,¹³⁵.

I. typographus and D. brevicomis are attracted to pheromone blends that have varied greatly in the ratio of components³⁷^,¹²⁰. At least in the case of I. typographus, this variation is due in part to host precursors¹⁶^,¹⁷^,⁷⁵ which are not under the control of the insect. Byers³⁰ suggested that myrcene and alpha-pinene were utilized by California bark beetles as precursors to pheromones because they were the most consistently present of the monoterpenes in host trees throughout a wide geographic range. Then, due to stabilizing-selection⁵⁰ (communication requires conforming individuals) and gene flow, the species would tend to equilibrate genetically in response to the availability of host monoterpenes, which were de-toxified and then used as pheromone components.

The rather large variation in monoterpene precursors would, however, produce selection pressures that would necessitate the evolution of some degree of behavioral tolerance of variations in pheromone quantity (and component ratios) in beetles utilizing myrcene and alpha-pinene. The inherent individual variation in component production would be further amplified proportionally by the monoterpene variation. Although behavioral tolerance of varying pheromone component ratios derived exclusively from host monoterpenes have not been tested, bark beetles seem rather more tolerant³⁷^,¹²⁰
than moths⁵⁰ of pheromone component ratios.

Bark beetles even appear somewhat tolerant of variations in the ratios of enantiomers of an attractive component. The percentages of (+)- and (-)-ipsdienol in the natural ipsdienol of individuals in a local population of I. pini (east Kootenay region, British Columbia) were found to vary. About half the individuals had nearly pure (-)-enantiomer, 20% had about 10% (+), while 5% had about 35% (+)¹³⁰. Over the geographic range of I. pini, the western population (British Columbia) averaged nearly 100% (-)-ipsdienol while the eastern population (New York) produced about 65% (+)- ipsdienol⁷⁹. The response of the western population is inhibited by the (+)-enantiomer, not because of an avoidance of eastern individuals but because of avoidance of its competitor I. paraconfusus which produces (+)- ipsdienol²⁰. The eastern population would not be attracted to western individuals because insufficient (+)-enantiomer is produced²⁰.

Few studies of bark beetles have measured the release of semiochemicals from natural sources due to the small amounts of semiochemical relative to contaminating host compounds. Browne et al.²⁷ quantified the release of attractive and inhibitory pheromone components of D. brevicomis boring in ponderosa pine by liquefaction of air. Since then porapak Q has been used as an adsorbent for attractive volatiles from T. piniperda⁴³^,⁴⁷^, ⁸⁰, I. typographus¹⁵^,¹²² and P. chalcographus⁴⁶. This type of data is necessary to formulate theoretical models of semiochemical interactions during colonization of the host tree.

Observations of the immediate attraction of D. frontalis to incipiently colonized trees under high beetle densities caused Vit� et al.¹⁴² to propose a "contact" pheromone whereby beetles would release pheromone upon landing on an attractive tree. While there have been no studies to determine the actual time of release, from an individual beetle's point of view it would seem better to release pheromone after acquiring an entrance tunnel with a female before committing resources (pheromone) to help in the colonization of the tree. Otherwise, a beetle might lose the advantage of arriving earlier and be outcompeted for access to females. In D. brevicomis males, which also contain frontalin and some inhibitors in the largest amounts upon landing (trans-verbenol, verbenone, and ipsdienol)⁴⁵, it would only make sense to an individual to release inhibitors from a place that is to be "defended against" competitors rather than just anywhere on the tree. This also applies for release of 3-methyl-2-buten-2-ol (MB) by male I. typographus who contain relatively large amounts upon landing¹⁷.

Olfactory perception of semiochemicals. The chirality of semiochemicals in bark beetles is important both in communication and in insect-plant interactions. The enantiomeric specificity is obtained by (1) the availability of chiral precursors of the host tree, (2) the chiral specificity of synthesizing enzymes, and (3) the enantiomer- specific acceptors on olfactory cells⁵⁵.

The electrophysiological response of an insect to chemicals can be measured using the electroantennogram (EAG) of the whole antenna or the single-cell technique which measures electrical responses of specific receptor cells. The antennal receptor cells each contain multiple acceptor sites that physically interact with the chemicals. Bark beetle olfactory cells on the antennae have been shown to be of several functional types, all of which generally are found in each species: (1) a "labelled-line" or highly specific receptor cell such as the ipsdienol-sensitive cells in I. paraconfusus and I. pini which are responsive only to one or the other enantiomer⁹⁴, (2) a pheromonal cell which is also responsive to some other synergists or inhibitors such as the frontalin cells of D. frontalis (the cells have at least two acceptor types specific for each enantiomer of frontalin)⁵⁴^,¹⁰², and (3) "generalist" receptor cells which respond to host monoterpenes as well as pheromones to some extent⁵⁴^,⁵⁵. However, the possibility that impurities in the monoterpenes account for the receptor activity has not been definitively ruled out.

The technique of "differential adaptation", in which a receptor cell (or antenna) is first adapted with a compound by saturation of specific acceptor sites and then exposed to a different volatile to see if an electrical response can be elicited, has been used to determine the types of semiochemical acceptors on a cell⁵⁵^,⁹⁹^,¹ ⁰⁰. By using this technique with single-cell recordings it has been shown that D. pseudotsugae has at least four olfactory cell types⁵⁵. Three types are each most sensitive to either MCH, seudenol, or frontalin, although they are all stimulated somewhat by all of these pheromone components⁵⁵. The fourth cell type is most sensitive to beetle-produced synergists and host attractants and is less sensitive to pheromone components⁵⁵. Acceptors are specific for one enantiomer of a chiral mixture¹⁰¹ and there may be either only one type of acceptor per cell (e.g. (+)- or (-)-ipsdienol in I. paraconfusus⁹⁴) or both types of chiral acceptors on the same cell (e.g. (+)- and (-)-frontalin in D. frontalis¹⁰² and D. pseudotsugae⁵⁵).

In I. paraconfusus, both sexes have equal receptor sensitivity (EAG) to natural pheromone and to (+)-ipsdienol⁸⁵. However, the males have been shown to be relatively less attracted by higher concentrations of synthetic pheromone components and were also not as likely to fly directly to the pheromone source as were females³³. Thus, the sexual differences in behavioral response appear to be the result of differences in central nervous system (CNS) integration. Both sexes of I. typographus have similar dose-response curves for cis-verbenol but some differences for MB⁵³. Similar to I. paraconfusus, males of I. typographus also are less directed to pheromone at the final landing than are females¹²¹; this behavioral difference could be due either to CNS differences between the sexes or to the lesser receptor sensitivity of males to MB.

Mustaparta et al.⁹⁴ have shown that eastern U.S. populations of I. pini have separate cells for the two ipsdienol enantiomers that are synergistic attractants⁷⁹. However, the exposure of the two cells to the enantiomers together did not synergistically increase the nerve impulse rate so it was concluded that synergism here acts at the CNS level⁹⁴. In contrast, Dickens et al.⁵⁵ have shown that the behavioral effects of MCH on D. pseudotsugae may be due in large part to peripheral receptors rather than as a result of CNS discrimination. At low concentrations MCH has acted as an attractant synergist¹¹² and certain olfactory cells have been found which are highly specific and sensitive to MCH in both sexes⁵⁵. The dose-response curve was judged to be relatively wide, suggesting a long-range orientation effect. However, at higher concentrations, MCH caused a decrease in spontaneous activity of other receptor cell types sensitive to other components⁵⁵. Thus, the behavioral inhibition of attraction by MCH¹¹²^,¹¹⁵ could be the result of CNS reception of the impulse frequency of the specific receptor type and also the peripheral inhibition of receptor activity⁵⁵. Whether this peripheral inhibition is a general phenomenon of semiochemical inhibitors is not known, although the ipsdienol inhibitors of Ips, mentioned above, appear to function only at the CNS level.

Of the three or four receptor types found in most species⁵⁴^,⁵⁵^,⁹⁴^,⁹⁵ it appears that individual receptors within a type also can vary in their response spectrum to various chemicals⁹⁴^,⁹⁵^,¹³⁷. In the few cases so far known, the host-responsive cells are present in both sexes but the host-selecting sex (males of Ips or females of Dendroctonus) is slightly more sensitive to plant compounds⁵³^-5⁵. However, the role of plant compounds in long-range orientation of Ips is not certain; while short-range behavior is probably influenced, it also is poorly understood. I. typographus is purported to be sensitive to exo-brevicomin since it increased response to the pheromone components¹³⁷, although it is not present in this species¹⁵^-1⁷. However, sympatric D. micans and/or Dryocoetus sp. produce exo- brevicomin and thus I. typographus could locate susceptible host trees by responding to species that colonize weakened hosts¹³⁷. This same phenomenon, on the behavioral level, was suggested for D. brevicomis and I. paraconfusus⁴⁰. However, the effects of "unnatural" semiochemicals on bark beetle receptors must be considered with caution. Lanne et al.⁸⁰ have shown that T. piniperda shows EAG responses to exo-brevicomin, ipsenol, and other compounds not known to be associated with the beetle or any competing species.

The difficulties with electrophysiological methods for unraveling ecological phenomena are two-fold: (1) the electrical responses may not be correlated to the relevant behavior and (2) the nerve impulse patterns vary even within a cell type⁹⁴, all of which are not located and tested, and the patterns are further integrated in the "black-box" CNS. It remains a great challenge to understand even incompletely the "cross-talk" interactions and "decoding" of nerve impulses from receptor cells at the peripheral level, a nearly impossible challenge when the CNS is considered.

Orientation to semiochemicals. Compared to moths⁵^,⁷², much less is known about orientation mechanisms of flying bark beetles in part due to their small size and difficulty of study in wind-tunnels⁵¹. A rotating wind-vane trap has been used to confirm the belief that bark beetles fly up-wind in response to a pheromone³⁶, however, little more is known of flight behavior. Byers³³ used grids of traps distant from a natural pheromone source and found that male I. paraconfusus do not fly directly to the source as females do, the reason being that males avoid patches densely colonized with potentially intense competition. This phenomenon has been observed in other polygynous species, e.g. I. typographus¹²⁰ and P. chalcographus⁴⁴. Sex-specific differences in D. brevicomis induced by trans-verbenol were observed when females were found to be inhibited from entering holes in artificial hosts releasing synthetic pheromone attractants while males were not. Sexual differences were not found in long-range orientation to pheromone component mixtures³², but single components (either exo-brevicomin or frontalin) caused the sexes to respond differently³⁵^,³⁷. The orientation of flying bark beetles is expected to use optomotor (visual) self-steered counterturning and anemotaxis in accordance with moth studies⁵^,⁷², although we have little evidence to support this belief.

Orientation by walking has been studied more often¹⁹^,²⁰^,³⁰^,³²^,³³^,⁴⁰^,⁴⁶^,⁴⁷ but with the objectives of observing attraction or inhibition by semiochemical blends. Recent studies have shown that bark beetles can orient in still air up a diffusion gradient¹⁴⁸. Although still air does not normally occur in nature this chemotactic mechanism may operate a few millimeters from an entrance hole. Host monoterpenes (camphene, alpha-pinene, �-pinene) of Norway spruce caused P. chalcographus to enter "beetle-sized" holes in cylindrical "tree-sized" traps more so than controls (both released pheromone components also)⁴⁴.

It was proposed that a long-range orientation function for a semiochemical was indicated if the compound had both a low threshold and a wide range of concentrations which elicited an electrophysiological response using EAG or single- cell recording. Conversely, a short-range function was indicated for a semiochemical if the electrophysiological activities were observed only with a relatively high threshold concentration and over a narrow range of concentrations⁵³. However, the evolution of an ability to detect a certain threshold concentration would depend on the absolute amounts available in nature, and thus a specific threshold (low or high) is not a valid indicator of the distance over which a semiochemical functions. Still, the breadth of the dose- response range (often shown to cover 5 orders of magnitude) does appear to be a valid indicator of a compound's role in distance orientation since a wider response to concentrations would be required over a wider range of distances from the source⁵³. Since (S)-(-)-cis-verbenol had an effect on receptor cell response over a wide range, it was suggested to function in long-range attraction in I. typographus, while MB had a narrow range (two orders of magnitude) and was thought to be a close-range signal⁵³. Field tests with various ratios of these two components were reported to support the concept¹²¹. However, when either component was released alone there was no significant difference in catch on traps 3 m from the source and few beetles were attracted¹²¹. Holding either component constant and varying the other caused a synergistic increase in attraction at the source¹²¹. Thus the theory of dose- response range width as an indicator of orientation range function needs more investigation - especially since the responses to MB were not tested at higher concentrations which might saturate the acceptors⁵³.

A possible complication when attempting to ascribe a short- or long-range function to a semiochemical is that receptor response to a wide range of dosages may also have evolved due to the natural variation in release rates of the semiochemical. MB synthesis is under endogenous control in male I. typographus and relatively large amounts are found in the gut (up to 4 ug/gut) relative to cis- verbenol¹⁶^,¹⁷. Thus it may be expected to vary less than cis-verbenol which is dependant on the quantity of (-)-alpha-pinene⁷⁵, a function of genetics and vigor of the tree¹⁷ in addition to the beetle's enzymes. The ability to respond to a wide range of concentrations would be required not only for orientation, where concentration decreases with distance⁹³, but also because aggregations of beetles vary in size and pheromone release rates³⁷.

A third possible factor complicating the interpretation of dose-response curves is that not all semiochemicals are equally volatile. MB is one of the most volatile of bark beetle pheromones and is found in one of the largest amounts per beetle (x=0.5 ug/male), while cis-verbenol is produced in small amounts (0.06 ug/male) and is less volatile¹⁷. E,Z-MD is one of the least volatile of bark beetle pheromone components (having a volatility comparable to moth pheromones) and is produced in the smallest of amounts (0.01 ug/male)⁴⁶. Whether this quantitative-volatility relationship holds more generally for bark beetle pheromones is not yet clear.