Brian Sullivan

Research Entomologist Insects, Diseases, and Invasive Plants (RWU 4552) USDA Forest Service


  • B.A. in Liberal Arts, St. John’s College (1990)
  • Ph.D. in Entomology (concentration in forest insects), Department of Entomology, University of Georgia (1997)

Professional Experience

  • Postdoctoral Associate, Department of Entomology, University of Georgia (1997-2000)
  • Research Entomologist, USDA Forest Service, SRS-4501/4552, (2000-present)

Research Interests

My research addresses the biology, systematics, chemical ecology, and management of native and invasive forest pest insects of North America.

Current Research

Southern Pine Beetle (SPB)

  • Manipulation with Pheromones and Other Odors. We discovered that a pheromone component produced by male southern pine beetles can increase attractiveness of the standard commercial SPB lure by 10-40 times. This improvement has greatly increased our capacity for detecting low level populations of this insect both within its current range and as it expands its range northward. The lure should improve forecasting of future outbreaks by providing more advanced warning of increasing beetle populations. Through tests of further adjustment to components of the lure and to methods used for deployment of lures and traps, we are maximizing the potency and utility of detection traps for SPB. Furthermore, our research has demonstrated that a number of volatile chemicals present in hardwood species of southern forests can inhibit attraction of SPB and thus may have utility as SPB repellants or tree protectants.
  • Understanding Variability in SPB (and bark beetle) Responses to Pheromones and Other Odors. Development of airborne chemicals as tools for manipulating behavior of SPB and other bark beetles has been challenged by conflicting test results, since not infrequently the same compound may appear to be attractive, repellant, or inactive in different tests. Our research has demonstrated that SPB responses to sources of pheromone are very different inside and outside of infested areas and that this variation is likely due to area-wide effects of a pheromone component released during beetle attack. Further investigation of these effects will improve existing capacity to manipulate and thereby manage SPB and likely other bark beetle pests of the US.

Chemical Ecology and Systematics of Bark Beetles of North and Central America

  • Bark beetle pest species new to science. Insect pests must first be identified correctly before successful management or fruitful research can occur. Our research with cooperators in Mexico has found that insects previously identified as SPB and agents of massive bark beetle epidemics in the Central American region are actually two different species--SPB and a second species new to science. This latter species, christened the “mesoamerican pine beetle” is nearly indistinguishable from SPB and appears to work in cooperation with SPB to kill trees. Outbreaks by the species pair appear to be more persistent and destructive than those by SPB alone. Recognition of this species complex may be critical to development of integrated pest management strategies for bark beetles in this impoverished part of the world; the discovery also brings to light a potential exotic threat to the US that was not previously known to exist. Additionally, our work with cooperators has demonstrated that western pine beetle (SPB’s western twin) has a different pheromone in the southwestern United States than in the Pacific coastal states; hence this very serious forestry pest of US, Canada, and Mexico apparently also consists of two distinct species.

Other Threats of Forest Trees of the Southern US

  • Baldcypress leafroller. We have identified the sex pheromone attractant of the baldcypress leafroller, a moth which causes defoliation and mortality to baldcypress in Louisiana wetland forests affected by rising water levels. Traps baited with this attractant can be used to monitor population levels and detect expansion of epidemics by this emerging, opportunistic pest. Additionally, the lure may potentially be deployed to disrupt mating by this species and thereby reduce the size of damaging populations.
  • Sirex noctilio. The Sirex woodwasp is a serious, exotic pest of pines native to Europe. It has invaded the northeastern United States and is expanding its range southward. We have been conducting research into identification of both the odor cues used by woodwasps to find and select host trees as well odors used by natural enemies to locate woodwasp prey. Synthetic mimics of these odors could potentially be used for wasp population monitoring as well as in novel management strategies.




steel anvil table
fiber optics lamp
Hewlett Packard 5890A gas chromatograph
muffler for stimulus delivery air
Pentium I personal computer
Peak Simple signal interface
Syntech CS-05 stimulus controller
splitter make-up pressure regulator
Syntech signal interface
transfer line thermal controller (Syntech TC-02)

antennal preparation
charcoal filter
cooling sponge for odt
indifferent electrode holder
odor delivery tube
pasteur pipette
recording electrode holder
transfer line (wrapped in fiberglass pipe insulation)


What is GC-EAD?

GC-EAD is coupled gas chromatography - electroantennographic detection. It is an analytical procedure that permits the rapid identification of compounds in complex mixtures that stimulate the olfactory sensilla of an insect. In other words, it can tell you what specific chemicals an insect can smell (and, to some degree, ones it can’t), and it can use odors derived directly from natural sources. This information can be used to discover potentially useful compounds—such as sex pheromones—that alter the behavior of insects.

In the 1950s, it was discovered that the voltage between the tip and the base of an insect’s antenna changed measurably when the antenna was exposed to odors of biological significance for the insect. This voltage is thought to represent the summed potentials of multiple responding olfactory neurons within the antenna, and the amplitude of the voltage roughly corresponds to an insect’s sensitivity to a particular compound. The voltage change does not reliably indicate whether a compound will influence the behavior of an insect or what the behavior might be, and it is quite common for compounds that produce strong antennal responses to have no observable behavioral effect. However, the presence of strong antennal responses (or responses to very low concentrations) indicates a greater probability that a compound will later be found to influence the behavior of an insect. Hence, antennal assays can assist in screening through the hundreds of odors found in the environment of an insect to permit the identification of those most likely to have behavioral activity. This can help in prioritizing compounds for behavioral tests and can greatly speed the identification of compounds that can potentially be used to modify insect behavior in beneficial ways.

Compounds can be exposed to the antenna after being purified or synthesized in a separate process; this procedure for manually exposing an antenna to compounds one at a time is known as an electroantennogram (EAG), and has received wide use in entomological studies since the late 1960s. The GC-EAD takes the EAG to a higher level of sophistication and utility by using the antenna of an insect as a detector for a capillary-column gas chromatograph.

Example of gas chromatograph detector output

The gas chromatograph (GC) is an apparatus used for separating and determining the identity and relative abundance of compounds present in complex mixtures of volatile and or semi-volatile compounds (Figure 1). Mixtures are flash-evaporated into a stream of inert gas moving through a long (typically 10-100 m), narrow (typically < 1mm) tube (called the “column”) lined with a semi-solid wax or polymer.

Figure 1. A basic gas chromatograph (GC). (A) A mixture of several compounds is injected into a heated chamber (the injector) where the sample vapor is mixed with a current of inert gas and is swept into the entrance of the column. (B) As the vapor is carried along in the flow of inert gas, compounds with greater affinity for the column lining will travel more slowly and separate from compounds with less affinity for the lining. (C) As compounds continue to move through the column, their separation from one another increases. (D) At the end of the column, each isolated compound exits (“elutes”) sequentially into a detector, which produces a voltage proportional to the amount of the compound passing into it. (E) The voltage output of the detector is plotted against time.

As compounds are carried along in the stream of flowing gas, some possess greater chemical affinity for the column’s lining than others and will travel more slowly. As a result, compounds will become separated as they travel within the column and will exit at different times (the “retention time” of the compound). Separation of complex mixtures by means of the differential partitioning of compounds between a “stationary” phase (in this case, the column lining) and a “mobile” phase (the inert gas) is a mechanism common to all forms of chromatography. In gas chromatography, the speed at which a compound travels is also dependent on the temperature of the column (compounds move faster at higher column temperatures). For this reason, the column is housed in an oven whose temperature can be precisely controlled. The effluent of the GC column is delivered to a detector which produces a DC voltage whose amplitude is proportional to the concentration of compound eluting from the column at that moment. The most common and most generally useful detector is the flame ionization detector (FID) which is sensitive to all organic molecules. The output of the detector is normally displayed on an X-Y graph with the Y axis representing detector voltage and the X-axis representing time. With the same column and GC operating conditions, a particular compound will always elute with the same retention time. Hence retention time is a diagnostic character that can be used to identify individual compounds in mixtures of unknowns.

Figure 2. A coupled gas chromatograph – electroantennographic detector (GC-EAD). (A) A mixture of several compounds is injected into a heated chamber (the injector) where the sample vapor is mixed with a current of inert gas and is swept into the entrance of the column. (B) As the vapor is carried along in the flow of inert gas, compounds with greater affinity for the column lining will travel more slowly and separate from compounds with less affinity for the lining. (C) The column effluent is split in two, with one half transmitted to a flame ionization detector (FID) and the other half carried to the insect antennal preparation. (D.1) The FID produces a voltage proportional to the amount of organic compound in the column effluent (D.2) Column effluent is mixed with purified air and blown over an insect's antennal attached to an amplifier. (E) The amplified electrical outputs of both the FID and antenna are plotted simultaneously

In GC-EAD, the effluent from the end of the column is split in two, with one portion of the effluent delivered to an FID and the other passed into a stream of purified air that is blown across an insect’s antenna (Figure 2). Electrodes (normally saline-filled glass needles) attached to the tip of the antenna and to the antenna’s base (or, alternatively, to the head or body of the insect) conduct DC voltages from the antenna to a high-impedance input amplifier which, in turn, feeds the signal to a graphical readout that simultaneously plots antennal voltage and FID voltage outputs against time (Figure 3). Synchronous voltage changes by both the FID and the antenna indicate olfactory sensitivity by the insect to the compound eluting at that particular retention time. The FID output can be used to confirm the presence, identity, and quantity of compounds exposed to the antenna while the antennal (EAD) output indicates presence/absence of olfactory sensitivity to eluting compounds and provides a relative measure of the intensity of olfactory stimulation.

Figure 3. Example of GC-EAD output. Extract from hindguts of female southern pine beetle, Dendroctonus frontalis, was analyzed on a GC-EAD fitted with an antenna from a male beetle. Three recognized compounds from the hindgut extracts are labeled. Several unstudied (unlabelled) compounds also produced antennal responses and are currently undergoing further study. It is noteworthy that, while trans-verbenol and myrtenol have demonstrated behavioral activity with male southern pine beetle (myrtenol is an inhibitor and trans-verbenol is an attractant synergist), cis-verbenol has not. The FID peaks of these three compounds were identified by matching their retention times to those of commercially-available versions of the same compounds injected into the GC. Identifications were confirmed by re-analyzing the extract with a coupled gas chromatograph-mass spectrometer (GC-MS) using the same column and GC operating parameters as used in the GC-EAD analysis.

Nifty Tools

Ultra-Fine Dissecting Probes

The dissecting probes above were fabricated for manipulating an antenna once it is mounted on the indifferent electrode. These tools are extremely helpful for inserting the end of a long filamentous antenna into the tip of the signal electrode. They are made by inserting the base of a minuten insect pin into the end of the wooden shaft of a cotton-tip applicator. A hole is initiated in the shaft with a heavier pin and the minuten is secured in place with glue (hot-glue was used for the probes shown here). The minuten can be bent into a variety of shapes for addressing specific needs.

Tool for Scoring (and Breaking) Pipette Tips

For the signal electrode to make “noise-free” contact with the tip of an antenna, it helps if the tapered end of the pipette is cut as cleanly as possible with no uneven or jagged edges. Crushing or breaking the pipette with a forceps almost never leaves a clean, circular opening. To prevent the glass from shattering, first score (“scratch”) the glass with the tool shown (above right), and then break the pipette tip off cleanly by gently grasping it with a forceps just beyond the scratches and twisting carefully (see Antennal Preparation). The scoring tool was constructed from fragments of a ceramic wafer (used for scoring GC capillary columns) that had been smashed with a hammer. Fragments with especially sharply-angled edges are inserted into a slit cut into one end of the wooden shaft of a cotton-tipped applicator and secured with a small amount of glue. The tool provides the precision and control needed for scoring the pipette tip at a specific, desired location

Storing Pulled Electrode Pipettes

The pipette storage device shown above was made by sticking a piece of adhesive-backed weather stripping to the bottom of a 10-cm diam. plastic petri dish and cutting a series of width-wise slits into the foam rubber. The pipettes are held securely within the slits, preventing damage to the sharpened ends. We use a forceps or a "triceps" for maneuvering the pipettes in and out of the slits. We got the idea from Eva Pettersson who used a strip of dental wax instead of the weather stripping.

Antennal Preparation

We have found that the preparation method detailed below works well for antennae of most insects we have examined, and it provides a high degree of reproducibility and a longer antennal life than other methods we have attempted so far. It has been used successfully with Coleoptera (6 species), Hymenoptera (7 species) and Lepidoptera (1 species). It is our standard method for making “first attempts” with a new insect species because of its reliability and simplicity.

Its major advantage is that the tip of the antenna does not need to be cut or punctured, reducing the number of steps required while simultaneously improving the usable life of the antenna. When surfactant is included in the saline (0.02% v/v Triton X-100 in Beadle-Ephrussi ringers solution) the signal/noise ratios obtained with this method are equal to those obtained using cut antennae. The basic procedure is outlined in the Syntech® manual (Electroantennography: A practical introduction 1998 version, p. 7) but we have elaborated on it.

The steps are illustrated here with a preparation of the clerid bark beetle predator, Thanasimus dubius:

  1. The insect's head is excised with a #11 scalpel blade on a pad of filter papers moistened with distilled water. The tip of one saline-filled electrode is broken at a point where its diameter is slightly less than the foramen (the opening between head and thorax) and then inserted into the foramen until the head is firmly secured.

  2. The pipette/head is attached to the indifferent electrode holder while a second saline-filled electrode is secured to the recording electrode holder positioned opposite.

  3. The tip of the recording electrode is broken off at a point where its internal diameter is slightly less than the maximum diameter of the antenna's (or the antennal club's) distal-most segment. The pipette break should be as smooth and symmetrical as possible. Clipping or crushing the pipette tip with a forceps invariably leaves a jagged edge or produces an opening of inappropriate size. For this reason we first gently scratch (“score”) the glass at the point where we want to break it using a piece of ceramic wafer secured in a wooden dowel (bottom left; also see Nifty Tools). Once scratch marks are visible on the glass, we very gently grasp the pipette tip with a sharp forceps just beyond the scratches and twist slowly in the direction opposite the scratches (bottom center). In most cases, the pipette will break cleanly leaving a smooth, circular opening (bottom right).

  4. The cut pipette tip is maneuvered with a micromanipulator until the saline just makes contact with the tip of the antenna (below). If the antenna is oriented perpendicularly or away from the tip, a small wire probe (see Nifty Tools) can be used to guide the tip of the antenna into the saline meniscus. Avoid any contact between the saline and the sides of the antenna. Once contact with the saline is achieved, the surface tension of the saline will maintain the antennal tip in contact with the electrode while the recording electrode is repositioned. (For most antennae, it is essential that the saline contain some wetting agent—such as Triton X-100 as described above—for the uncut antennal tip to adhere to the saline.)

  5. The recording electrode is moved away from the head to straighten the antenna and then moved in a roughly circular motion to gradually work the tip of the antenna into the pipette tip (bottom right). When finished, the antennal tip should effectively “plug” the opening of the pipette with its circumference, leaving a minimal amount of saline from the pipette in contact with the air (below).

  6. Below are some additional examples of this technique applied to antennae of other insects.

    A hymenopterous parasitoid of bark beetles, Coeloides pissodis (Braconidae)

    The southern pine beetle, Dendroctonus frontalis (Scolytidae). The antennal club is disk shaped, as is the case with many bark beetles. The opening of the pipette is made slightly smaller than the circumference of the club, and club is laid flat against the opening so that the most of the surface on one entire side is in contact with the saline. (This procedure was suggested to us by Dr. Qing-He Zhang of USDA-ARS Beltsville). In the specific case of D. frontalis, the preparation works better if no surfactant is added to the saline.

Mention of trade names is for information purposes only and does not imply an endorsement by the USDA.