Site, Stand, and Host Characteristics of Southern Pine Beetle Infestations
Jack E. Coster and Janet L. Searcy - editors
Southern Pine Beetle Handbook U.S.D.A. Combined Forest Pest Research and Development Program Technical Bulletin No. 1612
In 1974 the U.S. Department of Agriculture initiated the Combined Forest Pest Research and Development Program, an interagency effort that concentrated on the Douglas-fir tussock moth in the West, on the gypsy moth in the Northeast, and on the southern pine beetle in the South. The work reported in this publication was funded in whole or in part by the Program. This technical bulletin is one in a series on the southern pine beetle.
Gulf Coastal Plain
Ray R. Hicks, Jr., Kenneth G. Watterston, Jack E. Coster, and James E. Howard
The study area encompases Nacogdoches, Angelina, and San Augustine Counties in east Texas (Fig. 1), with a total area of 1.56 million acres, of which approximately 1 million acres are forested (Earles 1976). This land is part of the West Gulf Coastal Plains subregion. The major topographic features are low, rolling hills and upland flats, dissected by broad valleys and flood plains and ranging in elevation from 120 to 300 ft above sea level.
The forest is part of the "east Texas pineywoods" and is near the western limit of the contiguous range of southern pines. The principal tree species on upland sites include loblolly pine and shortleaf pine (Pinus echinata Mill.), with various mixtures of hardwoods. Hardwoods such as sweetgum (Liquidambar styraciflua L.) bottomland oaks (Quercus spp.), and hickories (Carya spp.) predominate in first bottoms, with baldcypress (Taxodium distichum [L.] Rich.) occurring along sloughs and streams. Loblolly pine may become a significant component of forests on second bottoms and natural levees.
|Figure 1. - Three-county study area in east Texas.|
Table 1. - Site and stand characteristics of attacked and baseline plots in natural, undisturbed stands-eastern Texas Coastal Plain
|Surface pH||log 1/[H+]||5.1||0.6||5.1||0.6|
|Surface soil depth||cm||43.3||29.4||47.6||31.5|
|Subsoil pH||log 1/[H+]||4.8||0.7||4.7||0.7|
|Site index at 50 yrs||ft||76.8||13.0||75.8||12.0|
|Avg. d.b.h. ("in" trees)||in||10.1||2.3||10.4||2.8|
|Avg. radial growth 0-5 yrs ago||mm||13.9||5.6||17.6||7.3|
|Avg. radial growth 6-10 yrs ago||mm||15.9||7.5||18.7||8.0|
*Mean and standard deviation for those plots with non-zero slope. Seventy-one percent of the attacked stands and 56 percent of the unattacked stands had no slope.
Upland forest types in the study area constitute the primary host types for SPB. Pine and pine-hardwood types cover about 85 percent of the forested portion of the study area.
The two largest ownership classes for forest land in the area are forest industry (about 44 percent) and nonindustrial private owners (about 45 percent). The remainder is split between U.S. Forest Service (approximately 11 percent) and other public owners (Earles 1976). The study area includes parts of three National Forests (Angelina, Sabine, and Davy Crockett).
SPB activity has historically been high in east Texas, with the greatest activity centered about 50 miles south of the study area. During the study (1975-77), an epidemic beetle population existed: the Texas Forest Service reported roughly 20,000 infestations in the area from 1974 to 1976.
Surveillance reports from the Texas Forest Service helped us locate beetle infestations. During 1975 and 1976, the infestations to be sampled were chosen in a random manner. But it became apparent that beetle activity was not uniform over the area, so the 1977 plots were allocated to insure collection of a representative sample.
The noninfested samples in this study are properly described as baseline plots, the purpose of which is to characterize the forest conditions in the area. Prior to 1977, we located plots in an unbiased manner so that each 5 x 5-mile grid square in the study area contained at least two plots. In the poststratification process implemented in 1977, additionally baseline plots were located in grid squares proportionate to the host-type forest land within the squares. Researchers collected 548 baseline plots, which was determined to be an adequate number to characterize the area (Bozeman 1977).
Results and Discussion
Characteristically, SPB-infested stands were located on flatter slopes than uninfested stands. Compared to the baseline plots, attacked plots were more heavily stocked and had a higher proportion of pines. Radial growth during the most recent 5-year interval was considerably less for the attacked plots, reflecting the more intense competition in such stands. Thus table 1 demonstrates that both site and stand factors leading to reduced growth and vigor are involved in predisposing trees to beetle attack.
Table 2. - Landform classification of attacked and baseline plots in natural, undisturbed stands-eastern Texas Coastal Plain
|Steep side slope||1.2||2.6|
Beetle infestations occurred at a higher-than-expected frequency on low-lying landforms such as flood plains, stream terraces, bays, and lower slopes. These landforms had a combined frequency of 20.2 percent for infested and 17.3 percent for baseline plots. Most attacked and baseline plots occurred on upland flats (56 percent and 42 percent, respectively). The higher uplands and sloping sites (perhaps drier sites) had a lower-than-expected attack rate. The combined percent of attacked plots on side slopes, steep side slopes, and ridges was 23.4 percent for attacked and 40.7 percent for unattacked stands.
Table 3. - Disturbances categories of attacked and baseline plots - eastern Texas Coastal Plain*
|No known disturbance||45||70|
|Logging activity within previous yr||8||6|
|Logging activity more than 1 yr ago||6||17|
|Ice and/or hail damage, severe-over one-half of stems affected||0||0|
|Ice and/or hail damage, light-less than one-half of stems affected||0||1|
|Fire within previous yr||1||1|
|Fire more than 1 yr ago||3||4|
|Lightening strike evident||32||1|
|Chem. brush control within previous yr||0||0|
|Chem. brush control more than 1 yr ago||0||0|
*A plot may have more than one disturbance.
Disturbances apparently triggered many infestations. Lightening strikes occurred at 32 percent of the infestations, while only 1 percent of the baseline plots showed this damage. Logging activity within the previous year represented the next most common stand disturbance. About 8 percent of the attacked plots had been disturbed by logging, while this phenomenon was recorded in only 6 percent of the baseline plots. Other disturbances were negligible for both attacked and baseline plots except that about 17 percent of the baseline plots showed evidence of logging more than 1 year before sampling, compared to 6 percent of attacked plots.
Table 1 shows that the average pine BA per acre is much greater for attacked plots. Table 4 shows that the distribution of BA also differs for the two stand categories. A definite positive shift in the total BA distribution can be seen for attacked plots. Sixteen percent of the attacked plots had BA less than 100 ft²/acre, while 50.2 percent of the baseline plots were within this range. Conversely, 58.8 percent of the attacked plots had BA exceeding 130 ft²/acre, as compared to only 21 percent for the baseline plots. Most of the infestations were small; 59.1 percent were less than 1/4 acre.
Table 4. - Basal area class (combined pine and hardwood) of attacked and baseline plots in natural, undisturbed stands-eastern Texas Coastal Plain
|Basal area class - (ft²/acre)||Attacked
|Average BA (ft²/acre)||143||109|
|Average percent pine||84||70|
Table 5. - Size and frequency distribution of infestations in natural, undisturbed stands-eastern Texas Coastal Plain (n=252)
|Acres - (nearest 0.25 acre)||Percent
In conclusion, it appears that about one-third of the infestations in eastern Texas were induced by disturbances such as lightening or logging damage that were unrelated to the presence of site, stand, or host characteristics. The remaining spots were associated mainly with wet, low-lying sites and with overstocked stands. The susceptibility of east Texas pine stands is, thus, determined by several predisposing factors; the degree of suspectibility depends on the number and level of these factors (Hicks et al. 1978b).
Timothy T. Ku, James M. Sweeney, and Victor B. Shelburne
Though SPB outbreaks historically have been rare in Arkansas, as outbreak began in 1969 in the southeast corner of the State and spread toward the northwest. At the peak of the epidemic, in 1977, SPB infestations occurred in 24 counties in south Arkansas (Fig. 2). Data collection began in 1975 and was concentrated in the southwest sector of the State; it followed the progress of the epidemic.
South Arkansas is characterized physiographically as Upper Gulf Coastal Plain. The soils are generally loamy in nature, and loblolly and shortleaf are the two native pine species. Shortleaf pine becomes dominant in the northern and western sections of the South. Most of the pine stands are natural in origin. The ownership is made up of many small, nonindustrial landowners and a few large industrial owners. This study includes all ownership types and represents a cross section of beetle-attacked areas rather than a concentration in a particular management type.
|Figure 2. - Study area in southern Arkansas.|
A preliminary analysis of 743 attacked plots and all control plots collected until June 1976 indicated that stand disturbance was significantly related to beetle attack. However, the data also suggested that undisturbed plots also suggested that undisturbed plots might be more indicative of inherent site and stand conditions that predispose an area to attack because of their larger area of infestation and greater number of trees killed. Therefore, data collection was redirected to SPB infestations showing no signs of disturbance. This stratification resulted in 230 natural, undisturbed plots on the Coastal Plain. This stratification was not extended to baseline plots collected from June 1976 to March 1977. These plots were established in the southwest corner of every township of the SPB-infested area.
Results and Discussion
The natural, undisturbed, attacked plots in Arkansas are younger, exhibit slower radial growth and thinner bark, and have a higher proportion of pine that baseline plots. The former also exhibit higher average BA/acre and occur more frequently in stands with high BA than in baseline plots.
Table 6. - Site and stand characteristics of attacked and baseline plots in natural, undisturbed stands-southern Arkansas Coastal Plain.
|Surface pH||log 1/ [H+]||5.3||0.4||5.2||0.4|
|Surface soil depth||cm||30.0||20.2||36.4||25.4|
|Subsoil pH||log 1/ [H+]||5.2||0.4||5.2||0.4|
|Site index at 50 yrs||ft||79.1||11.0||79.0||11.0|
|Avg. radial growth 0-5 yrs ago||mm||17.9||6.6||19.20||7.0|
|Avg. radial growth 6-10 yrs ago||mm||19.2||9.4||21.0|
* Mean and standard deviation for those plots with nonzero slope or aspect. Fifty-five percent of the attacked plots and 30 percent of the baseline plots had no slope.
Table 7. - Landform classification of attacked and baseline plots in natural, undisturbed stands-southern Arkansas Coastal Plain
|Steep side slope||0.4||0.9|
The distribution of plots in the landform classes indicates that more attacked plots occurred on upland flats, the most common landform type in the study area.
Table 8 depict the occurrence of disturbances in the initial 678 natural infested plots (not all plots are used in this comparison due to the sampling stratification) and natural baseline plots. Only 17 percent of the attacked plots are undisturbed, as opposed to 56 percent of the baseline plots. X² analysis on the frequency of disturbance showed that beetle infestation is significantly related to disturbance (P = 0.01). Lightening strikes and recent logging activity appear to be the most detrimental forms of disturbance, followed by ice damage and recent chemical brush control are not as important. Prescribed fire appears to have no relation to SPB attack. Nearly 80 percent of all southern Arkansas infestations were 1/4 acre of less in size.
Table 8. - Disturbance categories of attacked and baseline plots-southern Arkansas Coastal Plain*
|No known disturbance||17.0||56.0|
|Logging activity within previous yr||19.0||8.0|
|Logging activity more than 1 yr ago||25.0||21.0|
|Ice and/or hail damage, severe-over one-half of stems affected||4.0||4.0|
|Ice and/or hail damage, light-less than one-half of stems affected||20.0||8.0|
|Fire within previous yr||0.7||0.4|
|Fire more than 1 yr ago||1.4||3.0|
|Lightening strike evident||39.0||0.4|
|Chem. brush control within previous yr||2.5||0.0|
|Chem. brush control more than 1 yr ago||3.8||2.0|
*A plot may have more than one disturbance.
Table 9. - Size and frequency distribution of infestations in disturbed and undisturbed plots-southern Arkansas Coastal Plain (n = 973)
|Acres - (nearest 0.25 acre)||Percent |
|Lower limit||Upper limit|
Table 10. - Basal area class of attacked and baseline plots in natural, undisturbed stands-southern Arkansas Coastal Plain
|Basal area class - (ft²/acre)||Attacked
|Average BA in ft²/acre||128.4||106.6|
|Average percent pine||82.8||70.2|
Table 10 shows that attacked plots had higher basal areas. Eighty-five percent of baseline plots had BA < 130 ft²/acre, whereas only 56 percent of the attacked plots had BA < 130 ft²/acre.
Because both loblolly and shortleaf pines were present in the plots, a test was done to determine whether one species was more susceptible than the other. X² tests of SPB-attacked trees v. healthy trees for each species within all infested plots indicated that shortleaf pine was attacked more often than loblolly pine (P = 0.01). The same test was run on undisturbed plots; again shortleaf pine was attacked more often (P = 0.01).
In summary, the data indicate that while disturbance may indeed predispose a stand to attack, inherent stand conditions probably lead to larger and more devastating infestations. The lack of practical differences in site variables may be due to their interactions and resultant complexities. The high BA found in the undisturbed infested stands appears to be the major factor that causes competition, weakens the stand, and results in increased susceptibility to SPB attack. The slower growth may be merely an indicator of the competition.
Peter L. Lorio, Jr., and Robert A. Sommers
The 560,000 acres of commercial forest land included in the Kisatchie National Forest, in central Louisiana, represent a good cross section of pine sites on the Louisiana Upper Coastal Plain (Fig. 3). This forest and the data base associated with its management provided an excellent opportunity to study SPB site and stand relationships.
The Kisatchie timber management plan (FY 1972 through 1981, updated 8/28/78) describes the forest as about 90 percent pine and hardwood-pine types. Loblolly pine predominates, but shortleaf, longleaf (P. palustris Mill.), and slash pine (P. elliottii Engelm. var. elliotti) occupy considerable acreage. The forest is mostly middle-aged: 34 percent in age class 35 and below, 64 percent over 35, and 3 percent unclassified. Rotation age for the Yellow Pine Working Group (loblolly, slash, and shortleaf) is 60 years, and the prevailing site index at age 50 is 88 (ranging from 70 to 120). Rotation age for longleaf is 80 years and site index is 80 (range 70-100).
|Figure 3. - Study area on Kisatchie
National Forest in central Louisiana.
This study varied somewhat from the other ESPBRAP site/stand projects in that it assessed all SPB infestations within the national Forest study area and sampled a portion of them. Continuous Inventory of Stand Conditions (CISC) provided baseline data for the study (CISC is an automatic data processing system that reflects a continuously updated description of timber stands). We recorded pertinent information for each infestation from the appropriate compartment prescription summary sheets maintained by the National Forest. Data included forest type, stand condition class, site index, and age. Criteria for forest type are those of the Society of American Foresters. The primary factors considered in stand condition class are damage, quality, density, and age. Definitions are given in the Silviculture Practices Handbook (USDA Forest Service, FSH 2471.1-R8, Sept. 1974). Site index (Schumacher and Coile 1960) is reported by 10-ft classes, and age by year of stand establishment. Such data were recorded for all infestations, but plots were established only in infestations with five or more infested trees, and preferentially without disturbance.
We report data from two types of plots. First, data from single-point, 10 BA factor prism plots were collected and summarized as in the other Coordinated Regional Projects. This single-point plot served as the starting point for the second plot type, a hexagonal cluster of 10 sample points (Fig. 4). The single-point, 10 BA factor plot served as points 1 in the cluster plot. The cluster plots were established to get an even distribution over approximately 1 acre with points far enough apart to be reasonably independent (USDA Forest Service 1972). The central four points (1, 2, 5, and 9) represented the infestation origin; the outer six points represented either uninfested or subsequently infested sites; however, these sites are not comparable to control or baseline plots as reported in the other projects.
|Figure 4. - Plot design for study of SPB infestations;
point 1 is the location of the single-point plot (after
Forest Resources Inventory Work Plan, U.S. For.
Serv., South. For. Exp. Stn., New Orleans, LA.
Unpublished work plan.
The use of cluster plots provides more extensive sampling of SPB-infested stands than that possible with single-point plots. At each point in a cluster data were recorded for all pines >= 5.0 inches d.b.h. that fell within the limiting distance of 37.5 BA factor prism. Only BA was recorded for hardwoods.
Data from the cluster plots were summarized by a computer program specifically written for this study (Program written by Clifford A. Myers, Southern Forest Experiment Station, retired.). Summary data for each point included BA/acre of pine and hardwood, the number of pines per acre, and d.b.h., total height, and live crown ration of the pine of average BA. The summary program produces a card output that may be used to calculate current volume and to protect yield with program YLDTBL (Myers 1977).
Results and Discussion
Single-Point Prism Plots. - A total of 1,679 SPB infestation occurred on the study area during the 25-month study period, from June 1975 through June 1977. Of these 324 were sampled with 10-point cluster plots. Table 11 summarizes data from point 1 of 223 of the cluster plots that were established in natural, undisturbed stands. These 10 BA factor prism plots constitute the basic data for the Coordinated Regional Project.
Gently sloping sites with southerly aspect prevailed among all infestations. Soil texture averaged sandy loam in the surface and loam in the subsoil, with an especially wide range of subsoil textures represented. Soils were predominantly very strongly acid in both the surface and subsoil. Both the textural and soil reaction characteristics are representative of pine forests in the West Gulf Coastal Plain; no special relationship between SPB infestations and these site variables exists. Average surface soil depth of 42 cm (17 in) and loblolly pine site index of 96 ft, as well as the data on slope and soil texture indicate that infestations in natural, undisturbed stands were on productive pine sites.
Data in table 11 on age, bark thickness, and radial growth were limited to 931 trees purposely selected from the dominate stand for site index estimates. These trees averaged 16.1 inches d.b.h.
Infested stands averaged 77 percent pine, 142 ft² of total BA, and 39 SPB-killed trees. The diameter of the tree of average BA, including hardwoods, was 6.2 inches; this figure reflects the numerous small-diameter trees included in plot sampling. Stands averaged about 50 percent understory cover of vegetation not included in the prism plot.
Table 11. - Site and stand characteristics of attacked plots in natural, undisturbed stands-Kisatchie National Forest
|Surface pH||log 1/ [H+]||4.9||0.3|
|Surface soil depth||cm||41.5||20.1|
|Subsoil pH||log 1/ [H+]||4.8||0.3|
|Site index at 50 yrs||ft||95.5||11.8|
|Avg. radial growth 0-5 yrs ago||mm||14.6||5.3|
|Avg. radial growth 6-10 yrs ago||mm||15.3||5.8|
* Mean and standard deviation for those plots with nonzero slope and aspect. twenty-six percent of the stands had no slope.
Table 12. - Landform classification of attacked plots-Kisatchie National Forest
|Steep side slope||0.4||0.5|
The frequency distribution of landforms for plots in natural, undisturbed stands was very similar to that for all recorded infestations. Ridge was the landform most often reported. Many of the so-called ridges probably more closely resemble upland flats-nearly level or slightly concave areas on interstream divides, characterized by slow surface and internal water movement. Also, based on laboratory analyses, loamy and medium-textured soils may have been more frequently associated with infestations than field observations suggest (Figs. 5 and 6). Mapping unit codes 42200 and 44200 describe such sites. These mapping unit codes correspond to the high loblolly pine site index reported in table 11.
|Figure 5. - Distribution of 223 surface soil samples by soil textural classes. The (star symbol) marks average soil texture.||Figure 6. - Distribution of 223 subsoil samples by soil textural classes. The (star symbol) marks average soil texture.|
Table 13. - Disturbance categories of attacked plots-Kisatchie National Forest
|No known disturbance||41.1|
|Logging activity within previous yr||5.6|
|Logging activity more than 1 yr ago||0.4|
|Ice and/or hail damage, severe-over one-half of stems affected||0.8|
|Ice and/or hail damage, light-less than one-half of stems affected||0.7|
|Fire within previous yr||0.6|
|Fire more than 1 yr ago||0.7|
|Lightening strike evident||34.5|
|Chem. brush control within previous yr||0.4|
|Chem. brush control more than 1 yr ago||0.0|
Among all detected SPB infestations, 59 percent were associated with some kind of disturbance. Lightening was by far the most frequent disturbance encountered (35 percent). It was a significant factor associated with infestations over the entire study period, especially from midsummer through fall or early winter. At 12 percent, wind damage was the second most frequent disturbance. Most of the wind damage was associated with the Kisatchie Ranger District following a tornado in the spring of 1976.
Table 14. - Basal area class of attacked plots in natural, undisturbed stands-Kisatchie National Forest
|Basal area class - (ft²/acre)||Attacked|
|Average BA in ft²/acre||141.5|
|Average percent pine||77.3|
Table 15. - Size and frequency distribution of infestations in natural, undisturbed stands-Kisatchie National Forest (n = 223)
|Acres - (nearest 0.25 acre)||Percent |
More than half the infestations in natural, undisturbed stands had a total BA over 130 ft²/acre; 80 percent had 110 ft² or more. Stands with such stocking and the average characteristics reported in table 11 probably could support sizable infestations. However, most were quite small, less than 1/4 acre. Of the infestations, 80 percent were active when plots were established. But moderately low populations of SPB and a generally rapid salvage control program throughout the 25-month study period likely contributed to the predominance of small infestations.
Cluster (10 point) Prism Plots. - We found little variation among ranger Districts for tree and stand characteristics determined from the cluster plots in which acceptable plot trees were 5.0 inches d.b.h. and larger, and within the limiting distance of a 37.5 BA factor prism. Averages for characteristics of prism0selected trees (d.b.h.., height, and crown ratio) are for the tree of average BA. Based on these data the average SPB infestations in the Kisatchie National Forest was in pure pine type (78 percent pine) and immature sawtimber stand condition class (44 years old and 11.0 inches d.b.h.). Site index for loblolly pine was 93; BA/acre of the stand for this average SPB infestation exceeded the maximum leave BA suggested in the Technical Guide for Timber Marking³ (Section 2442.8, Kisatchie National Forest Supplement No. 46, November 1972, to the Forest Service Manual.) by 40 ft², if the hardwood basal area is included.
Average BA for the cluster plots was about 22 ft²/acre lower than for the single-point plots. Most of the difference was due to higher BA associated with single-point plot data-that is, in the immediate vicinity of the infestation origin. These data indicate a strong tendency for infestations to start in the more dense pockets of otherwise well-stocked pine stands.
In Louisiana, loblolly pine was the primary forest type attacked: 249 plots had 70 percent or more of their BA in loblolly pine. The principal apparent differences among the pine species were that the average infested stand of shortleaf was about 10 years older, had a lower total BA (106 ft²/acre v. 125 ft²), and a larger diameter (dominant stand) than loblolly (13.2 in v. 12.6 in). Infested longleaf and slash pine stands had considerably less hardwood then either loblolly or shortleaf, and the few infested slash pine stands were mostly young, well-stocked plantations.
Table 16. - Tree and stand characteristics of 10-point cluster plots (37.5 BA factor prism) in 318 southern pine beetle infestations-Kisatchie National Forest
|Ranger District||No. of
|Si²||BA (ft²/acre)||All pines||Dominant pine stand|
¹Six of the total (324) plots were deleted because of missing data. ²Height at age 50 for loblolly pine only.
Table 17. - Volume per acre of SPB-infested stands (loblolly or shortleaf pine > 70 percent of the pine component)-Kisatchie National Forest
|Average volume per acre¹|
¹Arithmetic averages of volumes computed for individual plots (trees > 5.0 in d.b.h.).
Based on the latest forest survey report for Louisiana (Murphy 1975), poletimber and sawtimber stands in the Kisatchie National forest averaged about 1,449 ft³/acre and 7,100 fbm/acre (Int. 1/4-inch rule). In applying Myers' program YLDTBL to 288 plots that consisted of 70 percent or more of loblolly or shortleaf pine, we found that SPB-infested stands had about 1.6 times these expected volumes. These volume figures are further evidence of the character of SPB-infested stands in West Gulf Coastal Plain forests. Although some variation existed among Ranger Districts, infested stands ranged from 2,200 to 3,100 ft³, 34 to 40 cords, and 10.9 to 15.8 M fbm/acre.
In summary, examination and analysis of the cluster plot data indicates that SPB infestations occurred in well-stocked to overstocked, natural stands of sawtimber-sized loblolly pine on productive sites. They tended to originate in the more heavily stocked portions of these stands. Hardwoods 5.0 inches d.b.h. or larger comprised about 22 percent of the stocking; and if this component were ignored, stocking would appear moderate when in reality it could be excessive.
Table 18. - Occurrence of SPB infestations by stand condition classes, all forest types-Kisatchie National Forest
|Stand condition class||Area¹||Infestations²|
|Damaged, sparse, & low quality||3.3||3.5|
|Damaged, sparse, & low quality||0.7||0.5|
|Seedling & sapling³||17.9||6.8|
¹Total area 522,704 acres.
²Total infestations 1679.
³Adequately and inadequately stocked.
Forest Resource Characteristics. - The Kisatchie National Forest is about 84 percent pine forest. SPB infestations occurred primarily in pine type -85 percent v. 10.5 percent in hardwood-pine, 3.3 percent in hardwood, and 1.2 percent in unclassified. Within the pine types, loblolly pine had a disproportionately high frequency of infestations (66 percent) and shortleaf a proportionate frequency (7 percent) in relation to area in those types. Longleaf and slash types both had lower infestations frequencies than might be expected based on area in those types. Virtually all of the infestations that occurred in hardwood-pine type were on lower slopes and stream bottoms where loblolly was the predominate pine species. Loblolly and shortleaf pine are both considered preferred hosts of SPB, but in this area loblolly may be the most abundant type as well as the most abundant type available to bark beetle populations.
Fifteen stand condition classes are recognized in the forest resource data. For our purposes, the 15 classes were condensed to 11 by combining the damage, sparse, and low-quality classes for the sawtimber and the poletimber size. Infestations were disproportionately distributed among stand condition classes (X² significant at .01 probability): a higher percentage of infestations occurred in sawtimber condition classes than expected based on the area they occupied. Immature and mature sawtimber stand condition classes had 68 percent of the infestations but occupied only 54 percent of the area. Immature sawtimber alone was associated with more than half the infestations.
Within pine management types (areas committed to producing a stand composed of 70 percent of a selected pine species), infestations occurred more frequently on the better sites as indicated by their distribution across CISC data site index classes: about 90 percent of the infestations occurred in site index classes 80 or higher. These observations agree with the results for site characteristics of the sample plots, as well as the high average stand volumes reported in table 17.
About 64 percent of the forest was in stand age classes 35 years and over; 83 percent of the infestations occurred in these age classes (Fig. 7). Distribution of infestations across 10-year age classes for pine management types showed that trees in age class 35 were most often infested. Further, for loblolly pine management type the area frequency distribution of age classes differed significantly from the SPB infestation distribution. Disproportionately more infestations occurred in age classes 35 years and older (X² significant at .01 probability), with a dramatic increase in infestations at age class 35 (Fig. 8).
|Figure 8. - Distribution of SPB infestations by
10-year age classes within pine management type.
|Figure 7. - Distribution of forest area and SPB infestations by stand age groups.|
In summary, forest type, stand condition class, site index, and age of stands reported in CISC data showed some consistent associations with SPB infestations. Loblolly pine type, immature and mature sawtimber, on sites with an index of 80 or higher, and in age classes 35 or older comprised the great majority of infestations.
Sincere appreciation is extended to field crew leaders Steven P. Weaver and John F. Kramer. Cooperation of the Kisatchie National Forest personnel, both in the Supervisor's Office and on the ranger Districts, is also greatly appreciated. The Office of the Regional Forester, Inventory and Plans, was most helpful in providing essential information on the CISC system and the data. Southeastern Area, State and Private Forestry, Forest Insect and Disease Management gave valuable assistance in locating infestations.
Southern Mississippi, Louisiana, and
Ronald J. Kushmaul and Michael D. Cain
Six timber corporations participated in this research effort. Their combined forest and landholdings included parts of 10 east Texas counties, 27 Louisiana parishes, and 19 southern Mississippi counties (Fig. 9). Approximately 2,235,000 acres comprised the study area.
During the research period (May 1975 to August 1977), corporate foresters, the Texas Forest Service, and the Louisiana Office of Forestry provided reports of SPB infestations. At the end of our research, the total number of reports received from each company's holdings by county or parish was multiplied by the average size of sampled spots on those holdings; that figure was then divided by acreage to obtain an intensity estimate for each company/county unit. Using these estimates, we grouped units with similar intensity levels and established the geographical distribution for five intensity levels (Fig. 10A - 10E). Since different intensity levels occurred within several Louisiana parishes (because of different company ownerships), those parishes are delineated more than once in figure 10.
|Figure 9. - Counties or parishes in the Louisiana, Mississippi, and east Texas study area (2,235,000 acres).||Figure 10A. - Geographical distribution of reported beetle activity - Intensity level 1 (0 to 0.149 infested acres/1,000 acres). Total acres 1,139,000. Longleaf, slash, and spruce pine, considered to be less susceptible in southeast Louisiana. Large acreages of longleaf and slash pine also occur in south central Louisiana and southern Mississippi.|
|Figure 10B. - Beetle activity at intensity level 2 (0.150 to 0.349 infested acres/1,000 acres). Total acres 459,000. Stand composition is, generally, mixed loblolly-shortleaf with some pure shortleaf stands.||Figure 10C. - Beetle activity at intensity level 3 (0.350 to 0.899 infested acres/1,000 acres). Total acres 273,000. Stands are mostly pure loblolly or loblolly-shortleaf, or infrequently pure shortleaf.|
|Figure 10D. - Beetle activity at intensity level 4 (0.900 to 3.999 infested acres/1,000 acres). Total acres 321,000. More pure loblolly stands occur here than at the lower infestation levels, but numerous mixed loblolly-shortleaf stands also exist. Also, there are a few pure slash stands.||Figure 10E. - Beetle activity at intensity level 5 (4.000 to 39.000 infested acres/1,000 acres). Total acres 43,000. Many overmature stands of loblolly mixed with shortleaf, while a few are loblolly mixed with slash or longleaf.|
In this study we obtained baseline data by measuring Continuous Forest Inventory (CFI) plots maintained by the six cooperating companies. Plots located in stands of pure longleaf or slash pine, and stands having only pines <5 inches d.b.h. were not sampled because they are normally less susceptible to beetle activity. We measured a 20-percent random subsample of approximately 1,800 CFI plots to establish a baseline data set representing the loblolly and shortleaf pine forests and mixed pine-hardwood forests in the study area.
Results and Discussion
Analysis of variance indicated that the means for the following variables differ significantly among the infested and baseline data set at least at the 0.05 level:
Slope, surface soil pH, subsoil pH, pine BA, total BA, stand understory, stand density, site index, average bark thickness-fissure, average bark thickness-ridge, average radial growth 0 to 5 years, and average radial growth 6 to 10 years.
Based on these variables, the characteristics of infested plots were:
Slope -- Flatter slopes had more
Soil pH -- pH of surface and subsoils was lower in infested plots.
Basal area -- Infested stands had higher pine BA as well as higher total BA.
Understory -- The percent area occupied by the understory was higher in infested plots.
Stand density -- Infested plots occurred in stands of higher density.
Site index -- Higher site indices were found for infested plots.
Bark thickness -- Trees in infested plots had thicker bark.
Radial growth rate -- Growth of infested stands had slowed during the last 10 years.
Table 19. - Landform classification of attacked and baseline plots in natural, undisturbed stands-Mississippi, Louisiana, Texas Coastal Plain
|Steep side slope||9.8||8.6|
Landform. - Infestations apparently occur more often in association with the less well-drained landforms. The relationship between individual landforms and beetle incidence was not significant (X² tests). When the cumulative frequencies of the various landforms are compared, however, the five flatter landforms have a disproportionate share of beetle spots. The lower average slope for infested plots compared to baseline (table 10) substantiates the observation that stands growing on flatter, less well-drained landforms were attacked more often. Lorio et al. (1972) related soil drainage conditions to susceptibility through the effect on root structure.
Table 20. - Site and stand characteristics of attacked and baseline plots in natural, undisturbed stands-Mississippi, Louisiana, Texas Coastal Plain
|Surface pH||log 1/ [H+]||4.6||0.4||4.8||0.4|
|Surface soil depth||cm||32.6||19.8||31.9||19.4|
|Subsoil pH||log 1/ [H+]||4.6||0.4||4.8||0.4|
|Site index at 50 yrs||ft||94.3||11.3||90.0||12.2|
|Avg. radial growth 0-5 yrs ago||mm||15.0||4.9||18.7||7.9|
|Avg. radial growth 6-10 yrs ago||mm||16.4||6.2||21.2||9.0|
*Mean and standard deviation for those plots with nonzero slope and aspect.
Table 21. - Size and frequency distribution of infestations-Mississippi, Louisiana, Texas Coastal Plain
|Acres (nearest 0.25 acre)||Percent of total|
|Lower limit||Upper limit||All plots
|Avg. spot size (acres)||1.7||2.4|
Disturbances. - The average size of infestations occurring in natural, undisturbed stands is greater than for infestations in general (table 21). Although stand origin (plantation v. natural) may also be a factor, these averages tend to substantiate earlier reports that disturbance-related infestations are usually small (Ku et al. 1976).
Comparisons of individual disturbances show some interesting differences between infested and baseline plots (table 22). First, recent logging activity occurs more frequently on infested spots than on baseline plots. Past logging disturbance, however, is much more frequently associated with baseline plots. This indicates that recent logging activity, perhaps because of the initial damage to trees and soil, is an "inducing" disturbance. Older logging seems to inhibit activity, though-perhaps because of the thinning effect. Second, recent fire slightly reduces beetle attacks. This could be because of improved soil moisture balance through the reduction in understory or other causes. Third, lightning is clearly associated with beetle activity, occurring much more often in infested stands than in stands in general. This finding agrees with earlier studies regarding the inducing effect of lightning on SPB attacks (Hodges and Pickard 1971, Ku et al. 1976). Finally, chemical brush control apparently has a long-term beneficial effect similar to that observed for logging.
Table 22. - Disturbance categories of attacked and baseline plots-Mississippi, Louisiana, Texas Coastal Plain*
|No known disturbance||33.6||31.3|
|Logging activity within previous yr||12.3||8.8|
|Logging activity more than 1 yr ago||8.6||21.8|
|Ice and/or hail damage, severe-over one-half of stems affected||0.4||0.0|
|Ice and/or hail damage, light-less than one-half of stems affected||2.0||9.9|
|Fire within previous yr||1.6||3.9|
|Fire more than 1 yr ago||16.0||14.8|
|Lightening strike evident||19.3||2.4|
|Chem. brush control within previous yr||0.8||0.0|
|Chem. brush control more than 1 yr ago||1.2||4.4|
*A plot may have more than one disturbance.
Basal area and site index.-Classed data for total BA (table 23) and site index (table 24) show that infested plot frequencies are skewed toward the higher values of both variables. Total BA and site index are significantly and directly correlated in both the infested and baseline data (table 25).
Since this correlation, especially for the pine component, is more pronounced in the infested data, beetle activity appears to be favored in stands where BA approaches the maximum potential of site quality. Therefore, stands with high BA may be more susceptible to attack during environmental stress conditions (e.g., drought) that temporarily reduce site quality.
Table 23. - Basal area class of attacked and baseline plots in natural, undisturbed stands-Mississippi, Louisiana, Texas Coastal Plain
|Basal area class - (ft²/acre)||Attacked
|Average BA in ft²/acre||139.0||96.5|
|Average percent pine||77.6||59.2|
The significantly lower average radial growth rate for dominant and codominant pines on infested plots (table 20) indicates that a decline in vigor, possibly caused by excessive competition, often preceded beetle attack. These data suggest the dynamic relationship between site and stand factors in predisposing beetle activity.
Cooperation from the following companies and State agencies was appreciated:
Boise Southern Co.
Louisiana Office of Forestry
Texas Forest Service
Table 24. - Site index class of attacked and baseline plots in natural, undisturbed stands-Mississippi, Louisiana, Texas Coastal Plain
(Base age 50 yrs)
|Average site index||94.3||90.0|
Table 25. - Correlation coefficient analysis of site index-basal area relationship-Mississippi, Louisiana, Texas Coastal Plain
|Variable pair||Level of significance¹|
|Pine BA with site index||0.003 S||0.440 NS|
|Harwood BA with site index||.004 S||.008 S|
|Total BA with site index||.001 S||.010 S|
¹The letter S denotes correlations significant at least at the 0.05 level.
Association of Annosus Root Rot with
Southern Pine Beetle Attacks
John M. Skelly, Samuel A. Alexander, and Roger S. Webb
The incidence of the root rot Heterobasidion annosum (Fr.) Bref. (Fomes annosus) in loblolly pine stands and its severity were measured in Coastal Plain forests in Virginia, Georgia, and Texas, to better understand the role of annosus root rot as an agent in predisposing loblolly pine and other pine species to attack by SPB.
We took the site, stand, and tree measurements described in other sections, and also excavated the entire root systems of all pine trees in the BA plots using a bulldozer. Unattacked plot data are of the control plot type. These plots were located about two chains from the attacked plot center. Each tree's root system was rated as to its overall size and appearance and a visual examination was made to determine if H. annosum occurred in the root system. If a tree had more than 1 percent of its root system affected by H. ammosum, detailed measurements were taken. Our field team measured the primary and secondary roots of all such trees, checking the number and length of affected and healthy roots and the base diameter of each root. Observers divided the measured root length into three symptomatic categories: (1) healthy, (2) resin-soaked, and (3) stringy. Categories 2 and 3 were considered symptomatic expressions of the presence of H. annosum. At periodic intervals, and whenever plots were being established in new geographical areas, roots with the three symptom categories were taken to the laboratory and the presence or absence of H. annosum was verified by selective isolation techniques.
Increment growth for the past 0-5 and 6-10 years was taken for each tree. Field teams also measured annual increment growth for all "in" pine. They removed a disk from each tree at breast height and made measurements at four points on the disk, the fastest and slowest growth rates, and the two adjacent sides.
Only those SPB sites infested less than 8 weeks were selected for plot establishment. Plots were established only in areas judged to be moderate- to high-hazard annosus sites. In moderate-hazard sites the surface soil contains 55 percent to 70 percent sand; high-hazard sites contain greater than 70 percent sand in the surface soil (Morris and Frazier 1966).
A total of 30 plots (15 attacked, 15 control) were established: 11 in Virginia (1100-series plots), 3 in Texas (1900-series plots), and 1 in Georgia (plot 1401).
Results and Discussion
The primary objectives of this study were to determine if an association between the occurrence of H. annosum and SPB infestations exist and if H. annosum predisposes pines to SPB attack by reducing radial growth. Plot locations were in Coastal Plain loblolly pine plantations and natural stands. A total of 335 trees in 30 plots were measured and their root systems excavated and rated, with 76 trees measured for H. annosum colonization.
In the attacked and control plots, field crews measured 52 and 29 trees, respectively, for H. annosum (more than 1 percent colonization). The average incidence of colonization for attacked and control plots was 37 and 45 percent, respectively; and for those trees with more than 1 percent of their root systems colonized, the incidence of colonization for all trees and those with more than 1 percent colonization was 21 and 13 percent, respectively. For the control plots the incidence of colonization for all trees and those with more than 1 percent colonization was 74 and 31 percent, respectively. Observers found conks in only 7 of the 51 plots.
Table 26 records the incidence and severity of H. annosum on moderate- to high-hazard annosus root rot sites in SPB-attacked and control plots. The overall incidence (at least one affected root per tree) of H. annosum in 92 trees on 9 attacked plots was 65 percent; in 98 trees on 9 control plots, the disease incidence was 73 percent. And 41 percent of the attacked plot trees had more than 1 percent of their root systems affected by H. annosum. In the control plots, 26 percent were so affected. Table 27 summarizes incidence and severity levels in natural stands. The incidence in 66 trees in 6 attacked plots was 24 percent, and in 79 trees in 6 control plots was 13 percent. None of the unattacked natural stands had trees with more than 1 percent of their roots colonized, and only 18 percent of the trees on SPB-attacked plots were so colonized. A comparison of plantations and natural stands indicates the same trends in colonization levels but with the higher amounts found in plantations.
Table 26. - Heterobasidion annosum colonization levels in attacked and control plots established on high-hazard H. annosum sites-Coastal Plain plantations
with > 1%
Table 27. - Heterobasidion annosum colonization levels in attacked and control plots established on high-hazard H. annosum sites-natural, undisturbed Coastal Plain stands
with > 1%
Table 28. - Average percent Heterobasidion annosum colonization of all trees in attacked and control plots-Coastal Plain plantations
|Plot||No. trees||% Colonization¹|
¹T-Test significance P = 0.05.
Table 28 summarizes the average percent of H. annosum colonization of all "in" pines for the attacked and control plots. We determined these percentages by measuring the primary and secondary roots of all trees with more than 1 percent of their root systems colonized, by assigning a value of 0.5 percent colonization for those trees colonized but having less than 1 percent colonization, and by assigning 0.0 percent colonization for those trees that were not colonized. The average H. annosum colonization for the attacked and control plots was 23.1 percent and 10.9 percent, respectively, significant at P=0.05. Table 29 summarizes the average percentage H. annosum colonization for the natural stands. The average percentage colonization for attacked and control plots was 12.0 percent and 0.1 percent, respectively. This is the same trend as found in the plantations but at a lower level. Variation in the natural stands exceeded that in the plantations.
Table 29. - Average percent Heterobasidion annosum colonization of all trees in attacked and control plots-natural, undisturbed Coastal Plain stands
|Plot||No. trees||% Colonization¹|
¹T-Test significance P = 0.028.
Table 30. - Number of roots in measured loblolly pine in attacked and control plots and the number and percent colonized with Heterobasidion annosum-Coastal Plain plantations
¹T-Test significance P = 0.03.
Table 30 summarizes the root colonization data for all measured (>1 percent H. annosum) root systems. The average total number of roots per tree measured and colonized with H. annosum per tree for the attacked plots was 32 and 14, and for the control plots 24 and 6, respectively. For attacked and control plots, the average percent colonization was 44 (0-58 percent) and 26 (0-48 percent), respectively. There was a significant difference (P = 0.03) between colonization levels with the higher H. annosum colonization in the SPB plots. In natural loblolly stands located in the Coastal Plain, the percentage H. annosum colonization was 12.5 and 0.5 for the SPB-attacked and control plots.
Table 31 records a summary of the total root length, both healthy and colonized with H. annosum, for SPB-infested and noninfested trees in the attacked plots. The percentage of H. annosum in the SPB-infested and noninfested trees in the attacked plots. The percentage of H. annosum in the SPB-infested and noninfested trees was 80 and 18, respectively, significantly different at P = 0.001. The percentage colonization in the secondary roots was significantly different (P = 0.01) with 44 and 13, respectively, for infested and noninfested trees.
Table 32 summarizes the total number of roots colonized with H. annosum and the number healthy for SPB-infested and noninfested trees in attacked plots. The percentage colonization was 86 and 32, respectively, for SPB-infested and noninfested trees. This difference was significant (P = 0.001).
Table 31. - Total root length infected by Heterobasidion annosum in measured loblolly pine in attacked and nonattacked plots-Coastal Plain plantations
|Plot||Tree no.||Total length||Total length
¹T-Test significance P = 0.001.
Table 32. - Total number of roots infected by Heterobasidion annosum in measured loblolly pine in attacked and nonattacked plots-Coastal Plain plantations
|Plot||Tree no.||Total length||Total length
¹T-Test significance P = 0.001
Table 33. - Average percent Heterobasidion annosum infection for SPB-infested and noninfested trees in attacked plots-Coastal Plain plantations
|Plot||No.trees||Average percent colonization¹|
|Infested trees||Noninfested trees|
¹T-Test significance P = 0.0001.
The average percentages of H. annosum colonization in SPB-infested and noninfested trees within the attacked plots were 54 percent and 11 percent, respectively (table 33). The difference between these mean percentages was significant (P - 0.001). The data indicate that within the immediate area attacked by SPB, H. annosum is associated significantly with those trees infested by the beetle. The conservative measurement of the level of H. annosum-colonized roots is indicated by the increased length of roots measured in trees with reduced levels of colonization. This would indicate more H. annosum in the root systems than was measured.
Annual radial growth, as measured on a dendrochronograph, was analyzed for each tree in the beetle-attacked control plots. The mean annual growth for the past 10 years was significantly different (P = 0.002) between the attacked and control plots. For 0-5 years and 6-10 years, SPB plots grew an average of 8 percent and 7 percent less, respectively, than the control plots. Mean colonizations for SPB and control plots were 23.1 percent and 10.9 percent, respectively (table 28).
Table 34. - Mean annual radial growth for each of the last 10 years for SPB-infested and noninfested trees located in SPB plots established on high-hazard (>70% sand) H. annosum sites-Coastal Plain plantations
|Avg. 0-5 yrs ago||1.24||1.72||-28|
|Avg. 0-10 yrs ago||1.49||1.69||-12|
¹T-Test significance P = 0.001.
Table 34 summarizes mean annual radial growths over the last 10 years for SPB-infested and uninfested trees. These plots were located on high-hazard annosus sites (>70 percent sans) in thinned loblolly pine plantations. The difference between the SPB-infested and noninfested was significant (P = 0.001). SPB-infested trees grew 28 percent and 12 percent less than the noninfested trees for 0-5 years and 6-10 years, respectively. These data establish the growth loss in SPB-attacked trees and the significantly higher levels of H. annosum present in the root system of these trees. This association strongly suggests that H. annosum was stressing those trees attacked by SPB, as expressed through the reduced radial growth.
Means and standard deviations for selected soil, plot, and tree variables in loblolly pine plantations and natural stands are summarized in tables 35 and 36. Comparing the average radial growth for 0-5 and 6-10 years for the SPB-attacked against that for unattacked plots in plantations suggests that reduced radial growth is associated with SPB-attacked trees. The average ages for the plantations were 30 to 27 years, respectively, for SPB and control plots. Soil data established the sites as high-hazard annosus root rot sites. Although plantations are generally uniform, the number of trees per acre was somewhat higher in the attacked plots.
Table 35. - Site and stand characteristics of attacked and baseline plots-Coastal Plain plantations
|Surface pH||log 1/ [H+]||4.8||0.5||4.7||0.8|
|Surface soil depth||cm||66.3||31.9||73.6||30.6|
|Subsoil pH||log 1/ [H+]||4.8||0.6||4.8||0.5|
|Site index at 50 yrs||ft||74.9||12.0||84.9||11.9|
|Avg. radial growth 0-5 yrs ago||mm||9.8||5.0||10.8||4.8|
|Avg. radial growth 6-10 yrs ago||mm||10.0||4.1||12.7||6.4|
|H. annosum root colonization||%||23.1||31.9||10.1||18.9|
*Mean and standard deviation for those plots with nonzero slope and aspect. Seventy-seven percent of the attacked stands had no slope.
Table 36. - Site and stand characteristics of attacked and baseline plots in natural, undisturbed Coastal Plain stands
|Surface pH||log 1/ [H+]||4.3||0.3||4.8||0.8|
|Surface soil depth||cm||61.5||23.3||56.7||23.2|
|Subsoil pH||log 1/ [H+]||4.6||0.5||4.8||0.5|
|Site index at 50 yrs||ft||76.6||10.8||72.6||32.7|
|Avg. radial growth 0-5 yrs ago||mm||7.6||7.0||7.0||5.2|
|Avg. radial growth 6-10 yrs ago||mm||9.8||11.3||9.7||8.9|
|H. annosum root colonization||%||12.5||27.9||0.5||1.9|
*Mean and standard deviation for those plots with nonzero slope and aspect. Eighty-three percent of the attacked stands had no slope.
The association of high H. annosum colonization levels with SPB infestation on high-hazard annosus root rot sites was clearly significant. This association between annosus root rot and occurrence of SPB infestations was greatest in thinned loblolly plantations. On low-hazard annosus root rot sites, however, this association was considerably weaker; and H. annosum appeared to play a much-reduced role regarding SPB attacks. Between SPB-infested and uninfested trees, the annual radial growth was significantly greater in uninfested trees, suggesting that the SPB-infested trees were less vigorous.
This study concludes that on deep, sandy soils (high-hazard annosus root rot sites) the incidence and/or severity of SPB infestations is associated with reduced growth rate (vigor) and this reduced vigor is due, in part at least, to the high disease levels of H. annosum in the root system. This conclusion points up the need for an applied disease and insect integrated pest management approach to controlling the SPB and other pests of the southern pine forest.
Developed by the University of Georgia Bugwood Network in cooperation with USDA Forest Service - Forest Health Protection, USDA APHIS PPQ, Georgia Forestry Commission, Texas Forest Service
and the Pests and Diseases Image Library - Australia
Last updated August 2018
www.barkbeetles.org version 2.0