Abstract - Sex pheromone traps are extremely efficient and effective for monitoring low density populations of spruce budworm, Choristoneura fumiferana Clem., when more conventional sampling techniques are too time consuming and expensive. They have now been deployed at more than 700 sites throughout North America for ten years as a method of monitoring low density populations to determine when more intensive larval sampling is warranted. On a local scale (forest management units) catches of over 100 moths/season are used as a threshold to trigger more intensive larval sampling. On a provincial and national scale, trap catches are converted from point sample data to contour maps by a geostatistical process called 'kriging'. The resultant maps are being compared with those from previous years by use of geographic information systems (GIS) to locate areas where populations are showing significant changes in density.
Key words - sex pheromone traps, monitoring, integrated pest management, geographic information systems, spruce budworm, Choristoneura fumiferana, Tortricidae, Lepidoptera
In contrast to most agricultural pests which are chronic problems recurring each year, most forest pests are cyclical, erupting into outbreaks at intervals separated by several years of vanishingly low densities. Most fluctuate on cycles of approximately ten years, but the spruce budworm, Choristoneura fumiferana Clem., fluctuates on an unusually long cycle of 35 to 40 years (Royama 1984, also see Figure 1). Up until now, increasing population densities leading to outbreaks have gone undetected until trees become visibly defoliated, by which time the only management option is to spray insecticides to keep the trees alive along with accelerated harvesting to salvage the trees before they are unusable. Early prediction of when and where outbreaks will occur would provide management with lead time to relocate harvesting operations or, possibly, to suppress the severity of an outbreak by slowing down population growth rates by the judicious use of insecticides.
Figure 1 Relationship between spruce budworm, C. fumiferana, larval densities and subsequent moth catches in pheromone baited traps at Black Sturgeon Lake in northwestern Ontario, Canada
Between outbreaks spruce budworm populations sink to extremely low densities - less than 1 larva/100 branches. At these densities larval sampling is impractical because it is far too time consuming and expensive. However, sex pheromone traps have considerable potential for monitoring changes in insect numbers at low population densities. In one location in northwestern Ontario, pheromone baited traps (using virgin females before synthetic pheromone became available) have been deployed every year since 1966. The trap catches together with larval sampling (Figure 1) show graphically the potential of pheromone traps for tracking changes in population density, particularly at low densities.
The sex pheromone of the spruce budworm has been identified as a 95:5 blend of E- and Z-11-tetradecenal (Sanders & Weatherston 1976; Silk et al. 1980). After extensive testing, the following combination of trap and lure is now recommended for monitoring changes in spruce budworm population densities throughout its range across North America.
The recommended trap is the Multi-pher I® (Sanders 1986, Jobin et al. 1993) manufactured by Le Groupe Biocontrôle (Ste-Foy, Québec), and available through a number of suppliers. The Unitrap® (International Pheromone Systems, Wirral, United Kingdom) performs equally well and is an acceptable substitute. Both types of trap have the capacity to hold several thousand moths without losing their effectiveness, thereby making them suitable for monitoring a wide range of population densities.
The necessary criteria for the lure formulation are protection of the synthetic pheromone from chemical degradation (usually oxidation, which is hastened by exposure to UV radiation), and release of the synthetic pheromone at a predetermined and relatively constant rate that spans the flight period of the spruce budworm plus a few weeks to permit the lures to be deployed in advance of the flight period. The chosen release rate is 100 ng/hr, slightly higher than the natural release rate of a female moth (Silk et al. 1980, Morse et al. 1982). This is sufficient to catch some males at very low densities, and yet it causes no aberrant behavior in the males, which is a possibility at excessively high release rates. The duration of release is specified to be at least eight weeks. This permits traps to be deployed during larval sampling, which may occur three to four weeks before moth flight, and yet it still spans the moth flight period (about 3 weeks in any one location). The lures currently recommended are Biolures® (Consep Membranes Inc., Bend, Oregon), with a loading of 2.8 mg of synthetic pheromone/ lure.
Each new batch of lures should be calibrated against the previous batch to ensure that they are of equal potency and correction factors should be applied where necessary (Sanders 1996). In Quebec, the Ministère des Forêts has used a different approach. There, correction factors are applied using the relationship between trap catch and density of overwintering second instar larvae (Boulet 1992).
In contrast to many of the traps used for the detection and timing of moth flight, the Uni-trap and Multi-pher traps contain no sticky surface, and if the moths are not immobilized in some way they will fly around inside the trap, damaging both themselves and other moths and making identification and counting difficult. The moths are therefore killed by an insecticide, for which registration is required. The insecticide selected for this purpose is the fumigant dichlorvos (DDVP). There are several products on the market with DDVP impregnated in plastic, but only one is registered for use in pheromone traps in Canada and the U.S.A.: namely Vaportape II® (Hercon Environmental Corp., Emigsville, Pennsylvania; Canadian Registration No. 21222; US Registration EPA No. 8730-32).
To ensure that trap catches are representative of budworm populations, certain protocols must be met. Traps are deployed in mature forest stands (a minimum of 10 ha in area) containing at least 50 percent white spruce, Picea glauca [Moench] Voss, and/or balsam fir, Abies balsamea [L.] Mill. Each trap is hung at eye level on a dead branch at least 50 cm from the stem of the tree, free from any obstruction that might prevent it from swinging freely in case the trap becomes snagged at an angle that allows moisture to enter and causes the moths to rot. As suggested by Jobin et al. (1993), hinged brackets can be fastened to trees in permanent sample plots. This ensures that the traps are hung in exactly the same position each year. Because trap catches are more variable at the edge of a stand, traps should be positioned at least 40 m from the edge.
One trap will provide almost as reliable an estimate as will a cluster of traps. However, three traps, arranged in an equilateral triangle with 40 m between traps is recommended. This allows for vandalism and damage to traps by bears.
Application in forest management
[top]
Index of population density
The correlation between trap catch and larval density was determined between 1989 and 1993 from about 40 sites in northwestern Ontario that cover a wide range of budworm densities. Pheromone traps were deployed each year throughout the moth flight period and the following winter branch samples were taken to determine the densities of the overwintering second instar larvae. The regression of larval counts against moth catch (Figure 2) shows that a catch of 100 moths corresponds to a density of 25 L2-larvae/10 m2 branch surface area, or about three larvae/branch, in the subsequent generation. This is approximately the threshold density at which sampling of the overwintering larvae becomes practical.
Therefore, in the boreal mixedwood stands of central Canada where these data were gathered, a trap catch of 100 can be used as a trigger to initiate a more intensive assessment of the situation. This can be done by a combination of more intensive pheromone trapping supplemented by larval sampling.
Trap catch is a reflection of the number of insects per unit area of forest. This of course varies with host tree density and size. In pure stands of mature host trees, such as white spruce stands in river bottoms in Alberta or balsam fir forests in Maine and the Maritime provinces of Canada, there will be far higher populations of spruce budworm per hectare than there will be in the mixedwood stands of Ontario that contain only 50 percent spruce and fir. Therefore, relationships between larval density and trap catch will have to be established for the major stand types involved in each region.
Figure 2 Relationship between spruce budworm larval densities and catches of male moths in Multi-pher sex pheromone traps as determined from 40 locations in Ontario during the period 1989 to 1993. The dashed lines denote the 95% confidence intervals. The dotted line shows that a moth catch of 100 corresponds to a larval density of about 25 overwintering larvae (L2)/10 m2 of tree foliage, or about 3 larvae/branch
Monitoring population changes
[top]
In order to monitor long-term trends in population fluctuations, including changes at low densities, traps are now deployed annually in approximately 700 sites throughout the range of the spruce budworm from Alberta to Newfoundland and across the six northern U.S. states (Figure 3). In order to analyze the data, however, it is first necessary to convert the point sample data provided by the traps into contour maps. A technique known as 'kriging' has been used for this. This technique, which was developed for mapping geological formations, has been applied to other point sample pest data (Liebhold et al. 1993). Briefly, kriging calculates a weighted average of trap catches for each pixel, based on the recorded catches from all locations falling within a defined radius. Details of the technique as it has been applied to the spruce budworm situation, together with details of the supporting software are described by Lyons et al. (1996). A sample map, spruce budworm moth density for 1994, is shown in Figure 4.
Figure 3 Map of eastern North America showing locations of spruce budworm sex pheromone trapping locations. The two areas of dense locations in New Brunswick are where intensive trapping was carried out to evaluate optimum distances between trapping locations
Figure 4 Map of spruce budworm moth catches for 1994 as determined from the trap catch data by the geostatistical technique of kriging
Figure 5 Maps showing areas where catches of male spruce budowrm moths exceeded 100 in a) 1991 and b) 1992, and by subtraction of a) from b), those areas c) where catches exceeded 100 in 1992 but not in 1991, which are areas of potentially new infestation.
These maps can then be used to determine changes in density from year to year, and can be superimposed on maps of cover type etc. to determine where the greatest risk of future damage is likely to occur. For instance, maps from successive years can be compared by geographic information systems (GIS) to show where trap catches have risen above the threshold level of 100 moths/trap during the past year as shown in Figure 5. More intensive sampling can then be carried out in these areas to delineate the boundaries of a potential outbreak, which provides forest managers with several years of lead time before defoliation first becomes visible.
Boulet B (1992) Surveillance des populations de la tordeuse des bourgeons de l'épinette: rétrospective 1986-1992, pp. A1-A4, in Insectes et maladies des arbres Québec 1992. Direction des communications et de l'éducation. Ministère des Forêts du Québec, Quebec
Jobin L, Coulombe C, Bernier-Cardou M (1993) Use of the Multi-pher trap to monitor spruce budworm populations. Forestry Canada, Laurentian Forestry Centre, Quebec, Inf. Rept. LAU-X-103E
Liebhold AM, Rossi RE, Kemp WP (1993) Geostatistics and geographic information systems in applied insect ecology. Annu Rev Entomol 38, 303-327
Lyons DB, Pierce B, Sanders CJ (1996) Data management system for the spruce budworm pheromone trapping network: user's guide. Natural Resources Canada, Canadian Forest Service-Sault Ste. Marie, NODA/NFP Technical Report, TR-38. in press
Morse D, Szittner R, Grant GG, Meighen EA (1982) Rate of pheromone release by individual spruce budworm, Choristoneura fumiferana, moths. J Ins Physiol 28, 863-866
Royama T (1984) Population dynamics of the spruce budworm Choristoneura fumiferana. Ecol Monogr 54, 429-462
Sanders CJ (1986) Evaluation of high-capacity, non-saturating sex pheromone traps for monitoring population densities of spruce budworm (Lepidoptera: Tortricidae). Can Entomol 118, 611-619
Sanders CJ (1996) Pheromone traps for predicting incipient spruce budworm outbreaks, Natural Resources Canada, Canadian Forest Service-Sault Ste. Marie, NODA/NFP Technical Report, TR-32. in press
Sanders CJ, Weatherston I (1976) Sex pheromone of the eastern spruce budworm: Optimum blend of trans and cis-11-tetradecenal. Can Entomol 108, 1285-1290
Silk PJ, Tan SH, Wiesner CJ, Ross RJ, Lonergan GC (1980) Sex pheromone chemistry of the eastern spruce budworm. Environ Entomol 9, 640-644