Discussion and Conclusions
The South-wide SPB Prediction System has proven to be a practical and relatively reliable means to forecast SPB activity early in the season. Three traps per county, parish, or ranger district appear sufficient to predict SPB infestation trends and probable severity at the local level. By averaging trap catch data for all counties and ranger districts monitored in a given year (20 to 100 traps per State), a reasonably reliable prediction of SPB trend and level for a given State can be achieved. There is a distinct relationship between the numbers of both SPB and clerids caught in early season traps and subsequent SPB infestation trends during the same year, providing further evidence that Thanasimus dubius populations are SPB-density dependent and are likely to be a major regulating factor in SPB population dynamics, as suggested by Moore (1972) and Turchin and others (1991, 1999).
Declining SPB populations tend to be more consistently predicted than are increasing populations, presumably because SPB and clerid populations are distributed across the pine-forested landscape more uniformly following an outbreak than they are at the beginning of the outbreak. Nevertheless, since its implementation in 1987, the prediction system has been used by State and Federal pest management specialists and foresters as a basis for scheduling aerial detection flights and identifying the need for suppression projects. Knowing when endemic SPB populations are about to explode or outbreak populations are soon to collapse is valuable information for more effectively managing this major forest pest.
Although 16 States currently participate in the system, levels of participation and accuracy vary from State to State. One State (Louisiana) prefers to use a single trap per parish and distribute available traps among more parishes. Interestingly, this approach doesn’t seem to have substantially affected the accuracy of State-level predictions (Table: Accuracy of the SPB Prediction System by state: 1987 to 2005, Table: Accuracy of the SPB Prediction System for 12 states from 1987 to 1998, Table: Accuracy of the SPB Prediction System for 12 states from 1999 to 2005). Presumably, surveying more locations or using more traps per location or both will lead to more reliable predictions, particularly in the early stages of a developing SPB outbreak when beetle populations are more scattered.
Continual validation using end-of-the-year detection records has enabled refinement of the system over the years. The prediction chart, in the Figure 1 on the right, has been modified since 1987 to better define the dividing lines between increasing or high, static or moderate, and declining or low infestations as more data points (both trap catch data and corresponding infestation trends and population levels) became available. Single sources of pheromone and turpentine have been used by all cooperators since 1999. Since 2006, the standard turpentine dispenser (amber bottle and wick) has been replaced by a polyethylene bag containing 200 gm of 75% alpha-pinene:25% beta-pinene (Synergy Semiochemicals Corp., Burnaby, British Columbia, Canada). This new dispenser is more practical for field application and has been found to be equally effective in eluting host volatiles and attracting SPB and clerids (Figure 2).
As with any prediction system, the SPB prediction protocol has strengths and weaknesses. The strengths include its simplicity, practicality, and relative accuracy. Data from a few strategically placed traps in early spring provide forest pest managers with insight into SPB infestation trends and levels for the remainder of the year at both local and large geographical scales. Currently, the standardized system is being implemented annually in 16 States, providing a consistent means to monitor SPB populations throughout the range of this pest in the Southern and Eastern United States. Limitations include the following:
- The process of installing and monitoring traps and counting beetles is tedious and time-consuming. Numerous field crews with varying levels of training and experience are involved in setting out and monitoring the traps and identifying the beetles captured, which influences the quality of the data.
- Pheromone traps placed too close to host trees may initiate infestations and inflate trap catches, leading to erroneous predictions, although such high trap catches are eliminated from the prediction data prior to analysis when infested trees are reported adjacent to a trap or trap catches exceed ca. 1,200 SPB per trap.
- Variations in climate among States means that trapping data from Northern States is often not available until early June, 1 or 2 months after the results from Gulf Coastal States are completed.
In certain years, atypical weather patterns may accelerate or delay emergence of overwintering SPB and clerid populations, leading to erroneous predictions. For example, in 1989, an SPB outbreak occurred in Texas but was not forecasted. In that year, beetles emerged in large numbers in late April and May, after survey traps had been removed from the field (Billings, unpublished data). Of course, SPB outbreaks may originate in counties or locations not monitored with pheromone traps, a potential problem in those States that survey a limited number of counties, (e.g., Mississippi, Alabama) or a single locality within each county or parish, (e.g., Louisiana) each year. Also, predictions made in the spring for increasing SPB populations will prove wrong if adverse summer temperatures or other unanticipated factors or both limit SPB population growth. And, finally, maintaining a reliable source of steam-distilled southern pine turpentine has been a problem over the years. Despite these limitations, the system has been deemed a success. This is reflected in the fact that Federal and State cooperators continue to use it.
Nevertheless, improvements in the SPB Prediction System are continually being sought. For example, monitoring SPB and clerid populations in the fall months for predictive purposes using the same protocol is currently being tested. Since 2003, the national forests in Mississippi have been conducting fall pheromone trapping during the secondary dispersal phase of SPB. These studies, though limited, have yielded relatively accurate predictions similar to those generated from the following spring survey (J. Meeker, personal communication). If effective, fall trapping would provide forest managers with additional lead time to prepare for peak infestations in the summer. The accuracy of forecasts based on fall insect numbers, however, may be dependent on weather conditions the following winter. Severe winter conditions such as freezes, prolonged cold temperatures, etc., could curtail the growth of SPB populations that were increasing in the fall, for example. In turn, SPB populations that are in the process of declining following a severe summer may be identified several months earlier if pheromone traps are deployed in the fall. Additional fall trials are needed across the Southeastern United States over multiple years to fully evaluate this approach. Indeed, a combination of spring and fall trapping may prove most effective for tracking seasonal and annual SPB population shifts.
Recent field research has demonstrated significantly enhanced attraction of SPB to trap sites additionally baited with the pheromone (+)endo-brevicomin (B. Sullivan, personal communication, East Texas Forest Entomology Seminar, Fall 2005). This finding offers a potentially new and alternative means of improving the precision, accuracy, and timeliness of the system. A bonus of the current SPB Prediction System is the wealth of historical trap-catch data from across the Southern United States over a 20-year period, which can help researchers better understand SPB and clerid dynamics (Tran and others 2007). Finally, incorporation of SPB predictions into area-wide hazard maps (Billings and Bryant 1983, Billings and others 1985, Billings and others 2008, Cook and others 2007) may lead to the development of more effective SPB risk maps. This would provide yet another tool for more effectively managing SPB.
Encyclopedia ID: p3297
