Douglas-fir Tussock Moth
Early Warning System
Factors Affecting the
Early Warning System
Plot Density and Distribution
In the 10 case studies, plot density based on initial year of defoliation averaged 1,648 acres per plot (range: 160 to 3,315 acres per plot; see list of outbreaks for details) for the seven cases in which a one-to three-year early warning was provided. Plot densities based on maximum defoliation for those seven outbreaks ranged from 733 to 20,561 acres per plot, with a mean of 7,695 acres per plot. Conversely, no early warning was provided in the two cases where plot density was low: no plots or 17,000 acres per plot for the area of initial defoliation, and 35,000 to 41,800 acres per plot for the total area defoliated. Further, in the latter two cases (Plumas/Lassen 1987-89, and southern Idaho 1990-92), the few plots present were clustered at one edge or within a limited section of the much larger areas that were subsequently defoliated.
The results of the case studies strongly suggest that areas to be monitored for DFTM should be supplied with a plot density of at least one plot per 3000 acres, based upon acres defoliated in the initial year. This density is about one plot per 5 sections (5 sq. mi) or about 8 plots per township. Additionally, the plots should be distributed proportionately across the area to be monitored, and not clustered along edges or in a limited sector.
Selection of Areas to be Monitored
Most plots are located in areas with a recorded history of DFTM outbreaks. During the past two decades, however, two outbreaks have occurred in part in areas with little or no recorded history of DFTM outbreaks (Plumas-Lassen and Giant Sequoia National Monument outbreaks). In addition, the amount of DFTM-susceptible host type has increased over the last several decades due to past management practices and fire suppression (Wickman 1992, Hessburg et al. 1994, Campbell et al. 1996). Thus, while plots in areas with a known history of DFTM outbreaks should be maintained, plots may also be warranted in other areas of potential susceptibility.
In general, plots should be evenly distributed throughout the host type at a density of about 1 plot per 3,000 acres. Plot density might be increased for areas with high relative value in terms of stakeholder concerns and management objectives - that is, specific areas where the short- or long-term effects of defoliation might lead managers to consider direct suppression. This criterion is relevant regardless of the recorded history of DFTM outbreak for a specific area, and could be used as an initial screen to help determine the distribution of early warning plots. One approach would be to allocate a higher density of early warning plots on those lands for which natural resource managers assign a higher priority for protection, while meeting the minimum density and distribution guidelines on other lands.
Moth Capture Thresholds
The trap threshold of 25 moths per trap was designed to detect increasing, but sub-outbreak, DFTM populations, thus triggering the need for follow-up ground sampling (Daterman et al. 1979, Shepherd et al. 1985). As shown in the geographic subregions figure, trap catch levels that signal an outbreak may vary for different outbreaks, but generally, traps in the outbreak area rise above the 25-moth threshold for one to three years prior to visible defoliation. A comparison of trap catch patterns within geographic subregions with the occurrence of specific outbreaks within those subregions demonstrates that increasing trap catches at the broader scale generally signal that an outbreak will soon occur somewhere within that subregion.
Average trap numbers across the broader scale may not be large, however. For example, in the Blue Mountains and central Oregon (subregions, Group B), average trap catches for all plots reached only about 15 moths per trap during periods of outbreak. In the vicinity of the defoliated area within the subregion, however, average trap captures were well above the 25-moth per trap threshold for up to three years prior to defoliation (figure 3, Northeastern Oregon - Malheur NF and Blue Mountains).
Due to within-plot and among-plot variation in trapped numbers of moths, the effective threshold will actually encompass a range of trap-catch levels, rather than the specific single value of 25 moths per trap. Shepherd et al. (1985), for example, reported that using six traps per plot would reflect a variation of plus or minus 30%, or a range of 17-33 moths per trap around an estimated threshold of 25 moths per trap. Because the early warning system uses 5 traps per plot, at the very least those plots averaging 17 moths per trap or more should be considered as potentially above the threshold.
Supplementary Plots
When faced with increasing trap catches, some managers have opted to install additional plots to temporarily supplement the information provided by the permanent plots. Although in some cases the Early Warning System triggered alerts up to three years prior to defoliation, in other situations the warning came only one year prior to defoliation. In the latter situation, a manager who waits one year pending the results from supplemental plots loses any timing benefits from the early warning. A more effective approach for improving early warning predictions of outbreaks would be to improve the density and distribution of permanent plots that are maintained annually.
Pheromone Components
The discovery of the new dienone pheromone component (Gries et al. 1997) raises the question of whether this new compound should be incorporated into the pheromone lures used in the Early Warning System. Addition of the dienone component would significantly increase attractiveness (Gries et al. 1997), even at the lower release rates calibrated for the early warning trap lures. However, Early Warning System lures were intended to have relatively low attractiveness so that traps would not become saturated at lower DFTM densities.
Significant changes in the attractiveness of the standard lure for the early warning monitoring traps would make meaningful comparisons with historic data difficult. Further, the case study results clearly show that the existing lure is effective for providing timely early warning of impending outbreaks when adequate numbers of plots are appropriately distributed across the areas selected for monitoring. Consequently, there appears to be no reason to incorporate the new compound into the monitoring trap lures.
Permanent Cocoon and Egg Mass Sampling Devices
Artificial shelters (often referred to as "cryptic shelters") have been developed as permanently installed sampling devices for collecting DFTM cocoons and egg masses (Dahlsten et al. 1985 1992, Sower et al. 1990). Late instar DFTM larvae readily spin cocoons and pupate in these shelters. The egg masses deposited by the flightless adult females can then be counted to measure population density and collected to determine egg mass viability.
The Early Warning System may be augmented in specific high value locations by use of these shelters. These passive sampling devices may provide site-specific indications of cocoon and egg mass densities as well as indications of associated natural enemy activity and other mortality factors. When maintained annually, the artificial shelters can give managers a timely, low-cost estimate of DFTM activity in high-value locations such as campgrounds or habitat for threatened or endangered species. If implemented on a plot, the artificial shelters may provide supplementary information on DFTM populations in the immediate area, in contrast to the information provided by the Early Warning System, which is representative of a much larger area.
Follow-up Ground Sampling
Follow-up ground sampling for DFTM pupae, cocoons, egg masses and/or larvae is a labor intensive and time consuming activity necessary for obtaining more accurate site-specific population density and natural mortality information, and for delineating the outbreak areas (areas of potential defoliation). Several ground-sampling techniques are available (Dahlsten et. al. 1992, Mason et al. 1993, Fettig et al. 2001).
Because tussock moth populations are highly aggregated (Shepherd et al. 1985, Mason 1996) and the pheromone traps may attract moths from distances of up to 4 miles (Daterman 1980, Shepherd et al. 1985), it is not uncommon for follow-up ground sampling to find low to very low DFTM population levels in the immediate area surrounding a Early Warning System plot with elevated trap catches. Consequently, to effectively assess the status of tussock moth populations, it is necessary to sample the general area (approximately 1- to 2-km radius) around plots and not just in the immediate area of the traps. It may also be appropriate to conduct initial ground sampling in areas of high management value, especially those near plots with elevated trap catches.