The Southern Pine Beetle

Chapter 5: Population Dynamics

Robert N. Coulson – Professor, Department of Entomology, Texas A&M. University, College Station, TX.

Introduction

Understanding the causes for changes in the distribution and abundance of southern pine beetles is prerequisite to developing an integrated management system for the pest. This knowledge of population dynamics can help predict where and when infestations will occur, how big they will become, and how long they will last. We can also use it to evaluate the probable effects of treatment tactics on SPB populations.

The goal of this chapter is to organize and interpret existing information on SPB populations with emphasis placed on the dynamic features. Implications of population dynamics of SPB to pest management decisionmaking will also be considered. There are five specific objectives of this review. The first is to define the basic population system in the individual trees and then to discuss how this system functions at the community (infestation) and ecosystem (forest) levels. The within-tree life processes will be used as elemental building blocks for understanding the community and ecosystem levels. Emphasis will be directed to the dynamic levels of the community and ecosystem. The second objective is to consider the role of the host in the population dynamics of SPB. The third objective is to consider the role of SPB in the population dynamics of the host. The fourth objective is to consider the role of weather in the beetle’s population dynamics. The fifth objective is to discuss utilization of information on population dynamics in integrated pest management decisionmaking.

SPB Populations Within Trees, Infestations, and Forests

The approach of utilizing a hierarchy of organization levels has been used to describe population dynamics of bark beetles (Coulson 1979). This approach permits definition of unique attributes associated with individual trees, infestations, and infested forests. The first level, the individual tree, includes information generally classed as natural history. By far the greatest volume of literature on SPB has been written at this level. It is at the second and third levels (infestation and forest, respectively) that actual dynamic features of population come into play.

SPB Populations in Individual Trees

Figure 5-1A – Spatial distribution of SPB life stages (life stage density v. the normalized infested bole height).	Figure 5-1B – Spatial distribution of SPB life stages (life stage density v. the normalized infested bole height).
Figure 5-1C – Spatial distribution of SPB life stages (life stage density v. the normalized infested bole height).	Figure 5-1D – Spatial distribution of SPB life stages (life stage density v. the normalized infested bole height).
The first level of organization includes the inseparable interrlationships between the beetle, associated microorganisms, and the host tree. Since the SPB’s life history is the starting point for a discussion of population dynamics, the following abstract scenario is provided here (see also Chapter 2). In the first stage of its life cycle, the adult SPB selects a suitable host tree, through either random or directed behavior (the exact mechanism is unknown). Colonization of the host is regulated by a blend of both insect-produced pheromones and host-produced attractants. Females initiate construction of egg galleries by boring into the inner	Figure 5-1E – Spatial distribution of SPB life stages (life stage density v. the normalized infested bole height).

bark region, where they are joined by males. Blue-staining fungi (Ceratocystis spp.) and other microorganisms are introduced at this time. Mating takes place within the galleries, and eggs are oviposited in niches at intervals along the lateral walls. Both males and females reemerge and are capable of attacking and colonizing new hosts. Eggs hatch shortly after oviposition, and the larvae excavate larval galleries at right angles to the egg galleries. There are four larval instars. The first three remain in the phloem region and the fourth migrates into the corky bark, where pupation and adult emergence take place. A large complex of natural enemies and competitors develops concurrently in the host. Figure 5-1 illustrates the general distribution of attacking adults (Coulson et al. 1976a), eggs (Foltz et al. 1976a), larvae, pupae, and emerging adults (Mayyasi et al. 1976a and b) in relation to the infested portion of the tree bole.

These spatial and temporal features of the beetle’s life cycle can be included by structuring the life cycle into a series of component processes: colonization (including attack, gallery construction, and oviposition), reemergence, brood survival, and emergence.

Colonization

The colonization process involves the location of a suitable host tree, identification of this tree for both SPB and its associates, aggregation (concentration) of sufficient numbers of individuals to overcome tree resistance mechanisms, inoculation of the tree with microorganisms, and establishment of an egg population. Successful colonization results in the death of the host (or a portion of it) and sets into motion a series of successional changes in the host tree that will subsequently have a pronounced influence on survival of the beetle population. For this reason, the colonization process will be discussed in considerable detail.

Attack. – The first phase of the attack process involves host selection by "pioneer beetles." These beetles must discriminate between host and nonhost species and between resistant and nonresistant hosts. The SPB’s mechanisms for host selection are poorly understood but apparently involve primary attractants, visual cues, and random searching. Host selection is undoubtedly guided by sophisticated behavioral mechanisms, because the requirement for identifying relatively rare susceptible hosts has important consequences for perpetuation of the insect. The host selection phase ends when adults successfully enter the phloem. At this time insect-produced pheromones and host-produced attractants are released.

The concentration phase follows host selection. Females respond to host trees marked by pheromones produced by the pioneering adults. These females initiate gallery construction, inoculate the tree with microorganisms, and produce additional pheromones. The pheromones, in combination with host attractants, stimulate the response of both sexes. After males arrive, mating, gallery elongation, and oviposition occur. These activities constitute the establishment phase, during which arrival of incoming beetles is apparently curtailed by production of inhibitor compounds. After oviposition is completed, the attacking adults reemerge. Figure 5-2 shows the general spatial and temporal sequence of SPB attack (ATK), based on measurements of within—tree populations. The four principal characteristics of the ATK processes are (1) pattern (the configuration of the curve of adult density v. tree height), (2) extent (the amount of host tree utilized), (3) density (the amplitude of the curve of attack density v. tree height), and (4) duration (the length of time involved) (Fargo et al. 1978, Coster et al. 1977). Each of these characteristics is influenced by the interaction of many variables. Arrival of attacking adults precedes actual entry into the tree by 1 to 2 days.

Figure 5-2 – Spatial and temporal distribution of attacking SPB adults (Y axis = adult density, X axis = height on infested bole, Z axis = normalized time). (From Fargo et al. 1978).

One noteworthy consequence of the sequence of attack is that the age distribution of the population varies in a predictable manner over the infested portion of the tree bole. The central section of the tree is attacked first and at higher density than observed above and below this area. Attack extends from the center section out to the end sections over a period of several days. Generally, attack density is lower in the end sections than in the center. Originally, investigators believed this lower population density at the extremes of the infestations was related to difference in host tree quality or quantity. However, Fargo et al. (1979) found no consistent relationship between tree physical characteristics (e.g., bark or phloem thickness) and beetle density. The observed spatial and temporal pattern, therefore, is likely a result of the sequence of production of behavioral chemicals.

During the concentration phase of colonization, attacks occur first at midbole and later spread toward the top and bottom of the tree. These attacks are guided by both the beetle- and host-produced compounds. After 2 to 3 days, following an increase in population density and establishment of galleries, inhibitor compounds are produced. The center portion of the tree, where colonization is advanced, becomes unattractive. Production of inhibitor compounds gradually spreads to the extremes of the tree over a period of time that follows the age distribution of the within-tree attacking adult population. If the adult population is available to attack a tree remained constant over the duration of the colonization process, the resulting pattern would be the same as observed in figure 5-2 (see Chapter 2). Production of pheromones by Ips. spp. and the black turpentine beetle may also be important in the sequence of colonization and in resource partitioning.

Attack density on individual trees is highly variable. Fargo et al. (1979) reported a range of 1 to 19 beetles/ 100 cm2, based on a study of 134 infested trees in east Texas. About three-fourths of the measurements taken were between 5 and 13 beetles/ 100 cm2 (fig. 5-3). The sex ratio of adults in successful attacks is essentially 1:1. The fate of the attacking adult population is influenced by many variables. The more prominent ones include meteorological conditions, predation, density, and quality of adults available for colonization, tree resistance, and physical attributes of the tree surface. The way these variables operate and interact through time and space is extremely important because for colonization to be successful, a population large enough to kill the host (or part of it) must be assembled. The time frame for the attack process ranges from 8 days to 6 weeks depending largely on seasonal conditions. Figure 5-2 is representative of a "mass attacked" tree. In this circumstance attacking adult populations are generally high and the process is completed in a short time.

Figure 5-3 – Frequency histogram of density of attacking adult SPB, based on estimates taken from 134 trees in east Texas.

Gallery construction and oviposition. – Gallery construction (GL) is an important component of the within-tree population system of SPB because there is a relationship between gallery length and oviposition. Considerable research attention has been given to this relationship. Knowledge of the fecundity of SPB is of obvious importance in understanding population dynamics. Furthermore, it is substantially less difficult to observe and measure gallery length than the number of eggs.

The general spatial and temporal pattern of GL is illustrated in figure 5-4 (Fargo et al. 1978). The GL process has the same attributes as described for ATK—pattern, extent, density, and duration. Gallery initiation generally occurs slightly below the center portion of the open bole. Gallery density is highest in this region and tapers gradually toward the top and abruptly toward the bottom of the infested bole (Fargo et al. 1978). Peak values for GL, based on a 3-year study in east Texas (Coulson et al. 1975a), ranged from 30 to 100 cm/ 100 cm² of bark (fig. 5-5).

Figure 5-4 – Spatial and temporal distribution of cumulative gallery length for the SPB. Y axis = cumulative gallery length density, X axis = height of infested bole, Z axis = normalized time.

Figure 5-5 – Frequency histogram of density of gallery length of the SPB, based on estimates taken from 59 trees in east Texas.

Recent studies have revealed that the relationship between egg populations and gallery construction is rather complex. Foltz et al. (1976a) demonstrated a linear relationship between eggs and gallery length: E (no. of eggs) = 1.59 X GL. This field of study, conducted in east Texas, utilized data collected throughout a summer season from several trees samples at various heights and aspects. F.P. Hain (personal communication) obtained similar results for populations in North Carolina. Other studies have demonstrated statistically significant departures from the results of Foltz et al. (1976a) and Hain. Working in Mississippi, Lashomb and Nebeker (1979) reported that in their field-collected samples many egg niches did not contain eggs. In their Georgia laboratory studies, Clarke, Webb, and Franklin (1979) observed that oviposition did not begin immediately as galleries were constructed. Regression analyses of Mississippi and Georgia findings produced different results from those obtained in Texas and North Carolina.

The effects of temperature, adult density, and seasonality on gallery construction and oviposition were investigated in lab studies by Wagner et al. (1980b). Interactions of these variable affected total gallery construction and number of eggs per female, duration of gallery construction and oviposition, and the shapes or configurations of the curves of cumulative gallery construction or eggs per female over time. A mathematical model incorporating these variables was developed. This model provides a realistic view of SPB reproduction.

The effects of temperature and adult density on GL and oviposition are less difficult to understand and explain than is seasonality. Variation in population behavior associated with season has been observed by a number of workers, e.g., Thatcher and Pickard (1964 and 1967), and Thatcher (1971) . Hedden and Billings (1977) reported seasonally related differences in fat content and size of adult beetles. Clarke et al. (1979) demonstrated that beetle size was associated with differences in fecundity. The relationships between beetle physical characteristics (size), energy reserves (fat), and fecundity are well documented. However, the importance of variations in beetle quality on population behavior in infestations and forests has not been examined in detail.

Resource utilization by SPB. – Obviously there are a number of important relationships between attacking adult and egg populations. One extremely critical relationship deals with the efficiency of host tree (resource) utilization. Each host has a finite quantity (or volume) of bark area suitable for brood development. The SPB has been shown to regulate egg populations through a density-dependent negative feedback mechanism that operates during colonization. The number of eggs per female varies as a function of the density of attacking adults (fig. 5-6). At low densities each female oviposits a larger complement of eggs. Conversely, at higher densities each female oviposits fewer eggs. Although information on within-tree SPB density has not been studied in detail, it is probably transmitted via acoustical signals. In the original description of resource utilization by Coulson et al. (1976a), egg populations were estimated using the constant E = 1.59 X GL from Foltz et al. (1976a). Figure 5-6 illustrates resource utilization based on both gallery length measurements and actual egg numbers obtained in laboratory studies (Wagner et al. 1980b). Although many variables have been shown to influence the relationship between egg populations and gallery length, estimates of egg populations based on a constant multiplier provide a reasonable approximation of the actual number.

Figure 5-6 – Relationship between gallery length (solid line) and eggs (dashed line) v. number of mating pairs of adult SPB, illustrating a decrease in eggs and gallery length with increasing adult density.

This mechanism of resource utilization permits efficient use of the host tree and prevents overpopulation that would result in mortality due to competition for available food material among brood life stages. Depletion of available host and food material is avoided by regulating the initial size of the egg population. Furthermore, rapid increases in populations are possible even though initial attacking adult numbers may be small. The insect, therefore, has the capacity to respond quickly to favorable host or weather conditions.

The resource utilization phenomenon has important implications in characterizing and forecasting population trends. A commonly used index of within-tree population trend is ratio of increase (Thatcher and Pickard 1964), which is defined as RI = (no. emerging adults) / (no. attacking adults / 2). Because of the resource utilization mechanism, the expected egg complement per female is higher at low attack density. Therefore, even with identical survival, the RI would be higher at low than at high attack density (Gagne et al. 1980b). RI has been proposed for use as an index of population "vigor" (Moore 1978), e.g., high ratios being indicative of vigorous populations and low ratios indicative of low vigor. But the index cannot be used for this purpose, because of the resource utilization phenomenon.

Reemergence

Reemergence is the process where adult beetles attack one host, lay eggs, and then exit to attack another host. This aspect of the natural history of SPB was recognized and reported by MacAndrews (1926 unpublished) in the first description of the life cycle of the insect. The significance of reemergence to population dynamics was identified by Franklin (1969 and 1970a), based on a study of infestation development in Georgia.

The spatial and temporal sequence of reemergence is illustrated in figure 5-7. The process has the same numerical features as reported for the components of the colonization process (pattern, extent, density, and duration). Reemergence begins at about the same time as peak arrival of attacking adults and continues for 10 to 14 days. The pattern and timing of reemergence follows the template established during oviposition. The number of beetles that reemerge ranges from 24 to 97 percent (Clark and Osgood 1964, Yu and Tsao 1967, Coulson et al. 1978 and 1980b, Cooper and Stephen 1978). The duration of the process over time is of survival value to the insect in that mortality agents affect only a portion of the population. Several important features of the behavior of reemerged adults have been identified from both field and laboratory studies. Females are capable of establishing two or three brood populations (Clark and Osgood 1964, Thatcher and Pickard 1964). Additional matings may be unnecessary for production of viable eggs in later attacks (Yu and Tsao 1967), although males are present in galleries constructed by reemerged females in the field. Reemerged adults produce pheromones that attract field populations (Coster 1970). Reemerged adults perceive and respond to pheromone and host attractants (Coulson et al. 1978).

Figure 5-7 – Spatial and temporal distribution of reemerged SPB adults. Y axis = reemerged adult density, X axis = height on infested bole, Z axis = normalized time. (From Coulson et al. 1978).

Reemergence is an important aspect of the beetle’s population dynamics and influences the pattern of infestation growth. Assuming that host trees are available in an area, there are five basic requirements for infestation growth (Coulson et al. 1978): (1) There must be enough SPB adults nearby and capable of attacking trees. (2) Host trees must be identified by the attacking population. (3) Initial host resistance to attack must be overcome. (4) Colonization and brood establishment must take place. (5) Local attractiveness must be maintained in the infestation.

Brood as well as reemerged adults contribute to infestation growth. The only biological attribute in which reemerged adults are known to differ from emerged brood adults is age. T.L. Wagner (personal communication) conducted extensive laboratory studies on reproduction of reemerged v. brood adults and found fecundity to be equivalent in both.

The process of reemergence adds another dimension to the resource utilization mechanism described earlier. Reemerged adults can oviposit all their eggs in one host or a portion of the complement in each of several hosts. When adult population density is low in an area, the former circumstance probably occurs. Conversely, when populations are high, the latter circumstance likely results. Distributing the eggs through multiple reemergences in response to variable density of adult populations is clearly a survival-enhancing mechanism for the beetle. Reemergence and resource utilization, therefore, are complementary processes. The fact that reemergence takes place where infestations develop enhances survival further in that the distance between hosts is small.

Survival of Within-Tree Brood Life Stages

In discussing the survival of within-tree populations of the beetle, we must consider the fates of eggs, larvae, pupae, and emerging adults. Survival of adult populations en route between trees will be discussed later. Generally, the factors that modify the distribution and abundance of populations include weather, food supply, intra- and interspecific competition, parasitization, predation, and genetic responses. These mortality processes may operate simultaneously and are exceedingly difficult to measure. The biotic mortality agents have been studied in considerable detail and are reviewed in Chapter 3.

Determining exactly what has killed a southern pine beetle is complicated by the fact that the insect is a native pest with a large complex of natural enemies. These mortality agents tend to concentrate in different portions of the infested tree and therefore have distinct and clumped distributions (Dixon and Payne 1979b). Population estimation necessitates the use of large sample sizes (Stephen and Taha 1976) that are generally impractical for field studies. Furthermore, quantitative definition of the population systems of the natural enemies is complicated by the rapid developmental time, probable interactions between species, and continuous growth pattern of the natural enemies and SPB.

In view of these complications, survivorship has been described as a general process for the within-tree population (Coulson et al. 1977). The spatial and temporal pattern of survivorship/100 eggs is illustrated in figure 5-8, which is based on an analysis of 149 trees sampled during 1972-74 in east Texas. The survivorship process has the same numerical attributes as colonization and reemergence—pattern, extent, density, and duration. One noteworthy feature illustrated in figure 5-8 is that average within-tree survival at various heights along the infested bole is virtually identical. However, survivorship varies among trees, host species, infestations, areas, and seasons. Also the mortality agents vary from one part of the tree to another. The net effect of the numerous biotic and abiotic mortality agents acting within the tree throughout development results in a uniform pattern of survivorship. The mortality agents therefore seem to have a similar net effect on the SPB, although different species operate in different regions of the same host trees.

Figure 5-8 – Spatial and temporal pattern of survivorship for within-tree SPB populations. Y axis = survivorship/100 eggs, X axis = normalized infested bole height, Z axis = normalized time.

The effect of the resource utilization mechanism on within-tree survival is demonstrated in figure 5-9, which illustrates survival of SPB/attacking adults over time and height on the infested bole. Although the beetle’s attack density is greater toward the center of the infested bole (fig. 5-2), the number of eggs/adult is less at midbole. The significant point is that the pattern of survival is the same for each section of the tree, while the actual number of beetles present varies. This observation further emphasizes the problem of using the ratio of increase to characterize population trends.

Figure 5-9 – Spatial and temporal pattern of generation survival for within-tree SPB populations. Y axis = survival of SPB/attacking adult, X axis = normalized infested bole height, Z axis = normalized time.

Figure 5-10 – Spatial and temporal distribution of emerged SPB adults. Y axis = emerged adult density, X axis = height on infested bole, Z axis = normalized time. (From Coulson et al. 1979b).

Emergence

Emergence is the final process in the beetle’s population system. Figure 5-10 illustrates the spatial and temporal pattern for the process. The same four numerical attributes occur — pattern, extent, density, and duration. Of the within-tree population processes, emergence has the highest degree of numerical variation. The observed spatial and temporal pattern of emergence reflects the net effects of all mortality agents acting on the population system.

The pattern for emergence is, again, a reflection of the original age distribution of the egg population. As with reemergence, emergence occurs in small daily increments over a period of ca. 14 to 28 days during the warmer months. This pattern likely has survival value for the species, as weather-related disasters would involve only a small part of the population. We also know that adults ready to emerge remain in the host during periods of inclement weather. Kinn (1978) described the diel emergence pattern for the SPB. Densities of emergent beetles vary widely. Estimates based on 134 trees ranged from 2 to 42 beetles/100 cm2 in east Texas (fig. 5-11). About three-fourths of the observations contained emergence densities in the range of 2 to 20 beetles/100 cm2. The influence of season on patterns of adult emergence in Texas has been investigated (Thatcher and Pickard 1967, Thatcher 1971, and Gagne et al. 1980a.). Salient features of these studies are discussed below in conjunction with the influence of weather on populations of the beetle.

Figure 5-11 – Frequency histogram of density of emerged SPB adults, based on estimates taken from 134 trees in east Texas.

SPB Population in Infestations

Populations of the beetle occur in clumps or infestations distributed throughout a forest. Infestations (= spots) are generally comprised of a number of trees, each containing beetles in one or more stages of development. Over the course of a summer, one spot can enlarge to include many infested trees. Spot growth is concentrated at one or more active fronts (or heads). This pattern of continuous spot growth is a unique feature of SPB populations.

Figure 5-12 illustrates a typical infestation. Although crown coloration is not always a good indicator of the stage of spot development, it does provide an acceptable means of illustrating the continuous growth pattern and variable age structure typical of infestations. In figure 5-12 the lightly faded green trees in the foreground have been colonized and reemergence of parent adults is occurring. The lightly faded red trees in the center contain developing brood life stages. Brood adults are emerging from the dark red trees in the background. The gray trees in the far background no longer contain beetles.

Figure 5-13 diagrammatically represents the sequence of development of a spot through the course of a summer season. The pattern of spot development is a function of the combined within-tree population processes in all trees in the infestation.

Figure 5-12 – Aerial view of SPB infestation (spot) illustrating trees in several age classes. Gray trees in background no longer contain beetles in any stage, red trees contain late developmental stages and beetles ready to emerge, lightly faded trees contain early developmental stages, and green trees in the foreground (at the edge of the spot) are being attacked.

Figure 5-13 – Spatial arrangement of attacked and unattacked trees in an SPB spot. The cylinders represent attacked trees and are proportional to the actual size of the trees in the spot. The ellipses represent peripheral unattacked trees. The two axes indicate the actual scale in meters.

Patterns of Continuous Growth

To track the continuous growth of a spot, researchers need both quantitative estimation procedures and knowledge of the spatial and temporal patterns of within-tree beetle population processes. For SPB there are now available several different sampling plans that have defined accuracy and precision (see Chapter 6).

In the original description of the processes of colonization (Fargo et al. 1978), reemergence (Coulson et al. 1978), survivorship (Coulson et al. 1977), and emergence (Coulson et al. 1979b), it was found that the distributions for each process could be averaged to produce a single curve (density over time) and still provide a realistic representation of a particular process in a tree.

Figure 5-14 illustrates an example of the within-tree population processes, expressed in two dimensions, for several trees sampled at different time during the development of a single spot. Density and duration are highly variable for each tree, but the general patterns for the processes are quite similar. Continuous estimates of attack, gallery construction, reemergence, survivorship, and emergence, together with the spatial and temporal location of both infested and uninfested trees, provide the basic information needed for analyzing and interpreting population dynamics of the beetle within spots (Coulson et al. 1980b).

Figure 5-14 – Arriving adults, centimeters of gallery construction, reemergence, and emergence, based on daily estimates taken from trees throughout the course of development of a spot.

Using the quantitative estimation procedures and knowledge of the beetle’s within-tree population processes, Coulson et al. (1980a) obtained daily estimates of populations from infestations. Figure 5-15 displays an example of the interrelationships between several of the beetle processes throughout the course of initiation, development, and termination of an SPB spot. Attack (A) is represented as a series of spikes occurring throughout spot development, as new trees are attacked. The attacking population is comprised of both reemerged (B) and emerged (C) adults. The two processes — reemergence and emergence (D) — were combined and termed "allocation" (Coulson et al. 1979b and 1980b). The two component processes of allocation vary throughout the summer. Furthermore, the total living brood within trees in a spot (E) influences the pattern of emergence. Numbers of infested trees and infested bark area are also illustrated in figure 5-15. The number of infested trees (F), total infested surface area (G), and average infested surface area per individual tree (H) have an important effect on the course and duration of spot development.

Figure 5-15 – Representative components of the population dynamics of the SPB, measured at the infestation level of organizational complexity. Measurements were made during the period of Julian dates 123 and 298 in a spot that occurred in east Texas during the summer of 1977. The beetle processes of attack, reemergence, and allocation are absolute numbers of beetles ÷ 103 brood alive within the trees ÷ 106. The host characteristics measured included the number of infested trees, the infested surface area (100 cm2) ÷ 103, and the mean infested surface area (100 cm2) per tree ÷ 102.

Analysis of continuous population data provides a means of interpreting processes unique to the infestations. Processes such as between-tree survival and allocation operate only at the spot level. These processes are important to understanding and predicting the distribution and abundance of the beetle. In addition, we can interpret the interrelationships between host and stand characteristics, weather, and population numbers using the continuous estimates.

Allocation of Adults

The distribution of beetle populations, whether in a tree, in a spot, or throughout the forest, is a function of the adult life stage. In any spot there are three categories of adults present: attacking parent adults, reemerging parent adults, and emerging brood adults. These classes can be catalogued further as within- or between-tree and of local or immigrant origin (Coulson et al. 1979b). Within-tree life processes that deal specifically with the adult life stage include colonization, reemergence, and emergence. A general depiction of these three processes in an individual tree is illustrated in figure 5-16. From this figure, it can be seen that a tree functions as a sink for beetle populations (attack), a sink and a source for beetles (attack and reemergence), or a source for beetles (reemergence and emergence). Within a spot there will be a mosaic of trees involved in each mode.

Figure 5-16 – Generalized plot illustrating the temporal relationship of the processes of attack, reemergence, and emergence for an infested tree. (From Fargo 1977).

Insight into the manner in which SPB infestations develop can be gained by considering the combined processes of reemergence and emergence from a single tree, i.e., allocation. On an individual tree, allocation (1) is continuous for each tree in the spot and bounded by the length of the within-tree life cycle, (2) is distinct for each tree, (3) is bimodal in intensity, and (4) its two components may operate together in concert or independently (Coulson et al. 1979b).

An individual spot contains a number of trees having different characteristics. The combined reemergence and emergence from all trees in the spot determines the number of adults available for attack and the resulting infestation pattern illustrated in figure 5-15. During periods of close synchrony between the components of allocation, a continuous supply of beetles is available for colonization. The rapid enlargement of SPB infestations in short periods of time is likely a result of close synchrony between reemergence and emergence, coupled with favorable weather and stand conditions.

Since the allocation process is distinct for each tree, mortality is independent for each tree in the spot. This fact can be of considerable importance in spot comprised of hosts of variable size and age distribution.

The bimodal form and independence of the components of the allocation process enhance SPB survival in that variables which influence one component will not necessarily affect the other. For example, inclement weather during reemergence may not persist or recur at time of emergence. Furthermore, the incremental pattern of reemergence and emergence, which results from the prolonged colonization period, ensures that only a small part of the allocating population will be exposed to short-term weather-related mortality.

The implications of genetic diversity resulting from allocation have not been examined. Colonization of a particular tree is accomplished by a blend of reemerged and emerged beetles. These beetles have different ages and origins from within the spot. Immigration may also be important. Reemerged females probably mate with different males than were present in the first trees to come under attack.

The allocation process has important implications to population dynamics. First, allocation is clearly an infestation-level process, with attributes that cannot be defined by examination of individual trees. Second, the concept of distinct generations, used to describe many insect populations, is not applicable to the SPB since (1) the colonizing population consists of beetles of different ages and origins, and (2) attacks on different trees usually occur over an extended period of time. Third, the age distribution of the within-tree population resulting from allocation is highly variable.

Between-Tree Survival Probabilities

Quantitative information on the fate of adults en route between trees in a spot or between spots in a forest has been extremely difficult to obtain. This information is of paramount importance to understanding population dynamics.

Historically, most of the research on populations of adults, under field conditions, has been conducted in association with studies of SPB behavior in response to pheromones and attractants. Conclusions from the studies were based on relative estimates of populations obtained from interception traps (as opposed to absolute estimates).

The early studies of Gara (1967) and Gara and Coster (1968) were among the first to examine patterns of spot development as related to the beetle’s response to behavioral chemicals. Spreading and collapse of infestations were found to be related to synchronization of beetle emergence (reemergence was not considered) with production of attractant compounds from nearby, newly attacked trees. The investigators also found that mass attack shifted from recently infested trees to vacant neighbors, and this behavior was influenced by proximity of host trees.

Subsequently, Coster et al. (1977a) and Coster and Johnson (1979) examined aggregation in response to attractive hosts and characterized the beetle’s flight behavior. Probability of attack on a tree in a spot relative to distance from an attractive host was defined mathematically (Johnson and Coster 1978). These studies were based on advanced understanding of chemical communication and provided an explanation for dispersal that was consistent with observed patterns of spot development and known responses to behavioral chemicals.

Information on the beetle’s flight behavior and quantitative estimates of daily populations involved in the processes of attack, reemergence, and emergence, taken together, provide the data needed to develop a model of allocation and a definition of between-tree survival for adults. Between-tree populations are composed of both reemerged and emerged adults during the time interval between leaving one host and successfully attacking another (entering its inner bark). Between-tree survival probability is the probability of an adult leaving one host and successfully attacking another. Allocation, in the present context, refers to the process wherein reemerged and emerged beetles are transferred from source to sink trees, i.e., from trees containing either reemerged or emerged beetles to trees being attacked.

Pope et al. (1980) and Coulson et al. (1980c) described two mathematical functions for transferring or allocating populations of SPB adults in a spot: a fixed probability function and a time- and temperature-dependent function. Both procedures produced similar results (fig. 5-17). The transfer function hypotheses were developed and tested against quantitative estimates of populations of adults in infestations (e.g., fig. 5-15). The basic problem that the transfer functions deal with is accounting for the distribution of source beetles (reemerged and emerged beetles) and sink beetles (attacking beetles) over a prescribed and realistic period of time. The accounting procedures can be checked for accuracy by comparing test results with actual measurements.

Figure 5-17 – Survival probability curves for emerged SPB adults held at five different constant temperatures (in °C) in the laboratory, illustrating the relationship between decreasing adult longevity and increasing temperature.

The assumptions of the fixed probability transfer function were that beetles could survive no longer than 6 days after leaving the host and that the proportion of surviving beetles on each day was a constant. The first assumption is probably reasonable; the second clearly is not. Nevertheless, the fixed probability transfer function provided a suitable description of actual field measurements.

The time/temperature-dependent transfer function utilized the same calculation procedure as the fixed probability procedure. However, adult longevity was based on temperature profiles defined from laboratory studies (fig. 5-17). Ambient conditions in the field were extrapolated from weather station data on temperature. Temperature had a significant effect on adult longevity. The time/temperature-dependent function provided a means of incorporating this important variable but, in doing so, added another degree of complexity to the model.

The probability of survival varies considerably over the course of development of a spot (fig. 5-18). The range covered an interval of ca. 0.75 to 0.05 survival.

Figure 5-18 – Survival probabilities for between-tree populations of SPB adults and the pattern of attack observed throughout the course of development of a spot in east Texas during 1977. The average maximum and minimum temperatures (C) are also included.

Figure 5-19 – Survival probabilities for between-tree SPB populations of emerged and reemerged adults throughout the course of development of a spot in east Texas during 1977.

By performing system perturbations on the time/temperature-dependent function, Pope et al. (1980) and Coulson et al. (1980c) demonstrated survival probabilities for segregated populations of reemerged and emerged beetles (fig. 5-19). The cyclic relationship between the survival probabilities for the two classes of beetles indicates that maintenance and perpetuation of a spot can be attributed to either reemerged or emerged beetles. In interpreting figure 5-19, it is important to recognize that the significance of increased or decreased survival probability is closely related to the absolute numbers of beetles available at any one time. For example, survival probability may be extremely high (day 249) but the number of beetles available small (day 249, fig. 5-15). The net result can be that the spot would not continue to enlarge.

Patterns of Spot Development in Relation to SPB Distribution and Abundance

In the previous sections I have discussed key elements of population dynamics of the southern pine beetle in spots. These elements included behavior of populations in response to pheromones and attractants; absolute population of attacking, reemerging, and emerging adults; the allocation process; and between-tree survival probability. Stand composition and geometry (the spatial arrangement of trees) together form the arena in which the life history processes take place. Therefore, it is not surprising that duration and intensity of spot growth are also linked, in part, to variables such as tree species, age, size, and density.

Characteristic pattern of SPB spot growth. – Development of beetle spots, by addition of newly attacked trees along one or more fronts throughout the course of a summer season, is a feature unique to SPB among the other Dendroctonus spp. There are at least three major reasons for this pattern of development. First, the reemergence process appears to be more pronounced with SPB than with other species. Reemergence takes place in the most recently attacked trees, i.e., adjacent to the next trees that will come under attack. These trees form the active front(s) (Coster, Hicks, and Watterston 1978). Second, developmental rate of SPB populations in the field is extremely rapid (30 to 50 days during summer months), compared to that of other Dendroctonus spp. Brood life stages, therefore, develop to the emerging adult stage and enter the colonization process in a short period of time. Synchrony between emerging and reemerging adults can be quickly achieved and maintained under favorable weather conditions. And the allocation process provides a continuous supply of adults for the purpose of colonization. Third, the fact that new trees are being colonized means that pheromones and attractants provide a continuous focal point for communication. Without the focal point, emerging adults, which are removed in space from the active front by ca. 10 to 20 m, would likely disperse and suffer substantially higher between-tree mortality than actually occurs (Gara 1967). Obviously beetle density; tree species composition; host susceptibility, suitability, and spatial distribution; weather; and myriad biotic mortality agents also influence the final success of populations in a spot. Nevertheless, the characteristic pattern of spot growth often results in rapid increase in beetle numbers and a corresponding high rate of tree mortality.

Dispersal patterns of reemerged and emerged beetle populations in spot growth are depicted in figure 5-20. This figure was developed from quantitative estimates of beetle populations taken from the spot portrayed in figure 5-13. These estimates were also the basis for calculating the survival probabilities illustrated in figures 5-18 and 5-19. In figure 5-20 the arrows point in the direction of spot growth. The size of the arrows is proportional to the number of beetles present throughout the course of spot development. The numbers on the figure are Julian dates at various stages in spot development. The allocation process continues throughout the entire period. At certain times either reemergence or emergence dominates in supplying beetles for colonization. Likewise survival probabilities for within-tree and between-tree populations and stand structure change throughout the course of spot development (figs. 5-18, 5-19, and 5-13). In the spot illustrated, about 50 trees were attacked during a period of around 100 days. This rate of attack is substantially less than often observed during epidemics, where 15 to 20 trees/day is not uncommon.

Figure 5-20 – Dispersal pattern followed by reemerged (solid line) and emerged (dashed line) SPB throughout the course of development of a spot in east Texas during 1977. The arrows are proportional in size to the number of beetles in either the reemerging or emerging mode. The numbers represent Julian dates in the development of the spot and range from 153 (June 2) to 249 (Sept. 6) in 1977.

SPB generations. – The life histories of many insect species can be summarized conveniently by generations and cohorts; and, therefore, considerable attention has been devoted to development of analytical procedures for the concepts, i.e., the life table approach to population analysis.

The within-tree population system for SPB can be considered as a cohort. But, as evidenced in figures 5-2, 5-4, 5-7, 5-8, and 5-10, the age distribution of the life stages has a very important temporal component. Therefore, analytical procedures based on the assumption of a stable age distribution for a population within a cohort in a generation do not strictly apply to SPB.

The production flow system approach, described by Coulson et al. (1976c), was based on point estimates of within-tree populations of SPB life stages. The life table approach was also used to evaluate the effects of a suppression tactic applied to within-tree life stages (Coulson et al. 1976d).

The pattern of continuous population growth in spots obliterates cohorts and generations of the beetle. This pattern results from the variable age distribution of beetle life stages in the attacked trees and the blending of populations in the allocation process. Therefore, the unique characteristics of SPB populations, which emerge at the spot level of organization, severely limit the applicability of analytical techniques developed for distinct cohorts and generations with stable age distributions.

At periodic intervals throughout the development of a spot there are surges or pulses in population abundance (fig. 5-21). In this figure population abundance for combined reemerged and emerged beetles is represented by the intensity of occurrence of the squares. The figure represents activity of populations of reemerged and emerged beetles throughout a spot’s development. Population surges occurred near Julian days 153, 173, 193, 214, and 249 (see fig. 5-20). Probably the result of interaction among many variables, the pulses reflect the influx of increased numbers of emerged beetles from previously attacked trees in the spot.

Figure 5-21 – Direction of dispersal followed by SPB adults throughout the course of development of a spot in east Texas during 1977. The boxes represent the direction (indicated by the line inside the box) of allocating SPB.

Indices of spot growth and decline. – Many biotic and abiotic variables interact to produce the wide range of growth patterns observed for SPB spots. It is not reasonable to expect that simple indices of population growth, measured at one point in time, will provide a consistent or reliable means for predicting spot growth. For this reason complex mathematical models are generally employed for making predictions (see Chapter 6). When food resources are not limiting, the fate of a particular spot becomes a function of within-tree beetle production (input to output ratios), between-tree survival probability, and developmental rate.

Several indices are suitable for portraying certain aspects of within-tree production (a quantity) and beetle productivity (a rate). Examples include emergence/attack, emergence/egg, reemergence/attack, emergence and reemergence/attack, and emergence and reemergence/attack/process timespan.

The main value of these indices is that they provide a simple representation of the net effect of the complex interactions of many variables. Gagne et al. (1980b) emphasized several limitations to the use of ratio estimators for characterizing population trends. Continuous population estimates for the processes of attack, gallery construction (oviposition), reemergence, survivorship, and emergence from each infested tree throughout the course of a spot’s development were used to calculate these ratios. A description of the field methods and analytical procedures needed to obtain the estimates for spots has been published (Coulson et al. 1979a and 1980b).

Examples of calculated values for the various indices of population trend are illustrated in table 5-1. Mean values were obtained from measurements taken throughout the course of development of three spots during 1977 in east Texas. Each index provides different information about the spots. None of the indices is suitable for predicting spot growth. The mean values of each index (table 5-1) for the infestations would probably not be statistically different.

Table 5-1. Summary statistics for time and survival components for within infestations of Dendroctonus frontalis.

The most commonly used index of population trend has been the ratio of emergence/attack (Thatcher and Pickard 1964, Coulson et al. 1976c and d and 1977, Moore 1978, and Gagne et al. 1980b). This index has also been termed "ratio of increase." The daily pattern for this index throughout the course of development of a spot is illustrated in figure 5-22. A value > 1.0 indicates that population output is increasing relative to the input. However, even when population numbers in a spot are large, a high value for the index may have little relationship to spot growth because of the variable nature of reemergence and between-tree survival of adults. This point is well illustrated by reference to the final days of development of the spot illustrated in figure 5-22. In this case the ratio was about 2.5, yet infestation growth ceased by Julian day 280.

Figure 5-22 – Daily ratios of beetle output per input in an SPB spot in east Texas during 1977.

The ratio of emergence/egg is a better index of within-tree brood survival than is ratio of increase. This index is substantially harder to obtain than ratio of increase because of the need to measure egg populations. However, the ratio is still preferable to ratio of increase since the number of eggs oviposited per female has been demonstrated to be a function of attack density (the resource utilization mechanism). This mechanism complicates interpretation of the significance of ratio increase. Mean values for brood survival from three spots are contained in table 5-1. In these infestations the ratio for each spot was about 0.19.

The ratio of reemergence/attack provides an index of the amount of redistribution of beetles that is taking place within a spot. For the spots in table 5-1, the ratio was about 0.60. This index has been shown to range from about 0.25 to 0.97 for D. frontalis (Coulson et al. 1978 and 1980b and Cooper and Stephen 1978).

The ratio (reemergence + emergence)/attack provides a measure of the actual beetle production per tree. This simple index incorporates both components of the allocation process (reemergence and emergence) and therefore is a better index of beetle production per tree than is ratio of increase. Mean values for the index from three infestations are contained in table 5-1. The index for each spot was ca. 2.4. The daily pattern for the index throughout one spot’s course of development is illustrated in figure 5-22. Values for the index would be extremely difficult to predict for infestation because of the resource utilization mechanism. However, attacks, reemergence, and emergence can be measured in the field (Coulson et al. 1980b).

The index [(reemergence + emergence)/attack]/process timespan is probably the best single measure of beetle input-output per tree. This index represents the rate at which beetle production per tree is taking place. The main problem with the index centers on measuring the process time components in the field. Mizell and Nebeker (1978) described a procedure for estimating developmental time using field temperature measurements. The pattern observed for the index throughout the course of development of an infestation is also illustrated in figure 5-22.

Another seemingly useful measure of spot growth and decline is the number of living brood within the tree comprising the spot through time (fig. 5-23). Again, this index incorporates the interaction of many variables and ultimately reflects the number of beetles emerging into the spot. It is interesting to note that in this figure the number of brood alive within the three spots illustrated was normally the same at ca. Julian day 160, yet each spot declined at different times and at different rates. This observation simply illustrates the futility of attempting to utilize simple indices as a means of predicting population trends in spots.

Figure 5-23 – Daily estimates of the total within-tree brood alive in three SPB spots in east Texas during 1977. These estimates span the duration the spots were active. Note that about day 275 all spots contained nominally the same number of living brood stages, yet the spots each became inactive at different times.

Each of the five spot-growth indices described above provides information about the status of SPB populations in a spot. But none of the indices is suitable for predicting the future pattern of spot growth. Also, there are many other possible indices besides those presented above. Coulson et al. (1979c) described 12 different measurements that characterize spot growth patterns. All the indices require a substantial direct measurement effort in the field. For instances where these indices are used to characterize populations, calculated values will likely be generated as output from a model rather than measured directly in field studies. This output can then be used to define further basic principles of population dynamics, evaluate the efficacy of treatment tactics, predict timber mortality resulting from SPB infestations, etc.

SPB Populations in Forests

At the ecosystem level, emphasis on the beetle’s population dynamics centers on the distribution, number, and size of spots. The same principles that govern SPB distribution and abundance on trees and in spots operate at the ecosystem level. Generally the main variables that limit the distribution and abundance of SPB at this level are climate and the availability of susceptible and suitable host material. The effects of weather and climate on populations of SPB will be discussed below. The distribution and abundance of susceptible host type are influenced by a number of variables associated with tree species, soil and site conditions, and cultural practices. These variables have been discussed by Hicks in Chapter 4. Salient features of the interrelationship between population dynamics of SPB and tree species and soil, site, and stand conditions will be discussed below.

Host’s Role in SPB Population Dynamics

Host susceptibility (= degree of resistance) to insect colonization and host suitability for brood development are subjects of considerable importance in understanding SPB population dynamics. The approach of employing trees, spots, and forests used previously is also useful for organizing concepts of host susceptibility and suitability. Again, the beetles interact with individual trees, stands, and forests.

Susceptibility and Suitability of Individual Trees

The concepts of host susceptibility to colonization and suitability for development are tied to inseparable interrelationships between the beetle, associated microorganisms, and physical and chemical characteristics of the host. Colonization initiates a series of events that eventually lead to the death and subsequent degradation of the host. The first step in the sequence involves overcoming tree resistance mechanisms. If this phase is successful, the tree, or a portion of it, will die. Without tree death, brood stages will not develop.

The blue-stain fungus, Ceratocystis sp., is considered to be the principal tree-killing agent, although many other microorganisms (bacteria, yeast, and other fungi) have been identified from the beetle’s mycangium and body surface. Many details relating to beetle-fungi-host interactions have been described (see Chapters 2 and 4), but there has been no attempt to develop a comprehensive interpretation of the SPB’s role in the scenario. Safranyik, Shrimpton, and Whitney (1975) have provided a conceptual statement for the mountain pine beetle (D. ponderosae Hopk.) in lodgepole pine (Pinus contorta Doug.) that is generally applicable to the SPB. The following description is based largely on their interpretation.

Attacking adults arriving on the host chew into the phloem and thereby begin to inoculate the host with blue-staining fungi. The spores germinate rapidly and penetrate the phloem and xylem as the beetles enlarge their galleries. Primary resin seeping from damaged resin ducts slows the attacking beetles but the production of secondary resin by ray cells is apparently the process that can prevent establishment. If the phloem and sapwood next to the wound become saturated with resin, the beetle will be killed (resinosis) or repelled (pitched out) and the fungus will die. This circumstance occurs when the tree response, in the form of resin production, is rapid and massive.

Longleaf and slash pines characteristically produce greater quantities of resin than loblolly or shortleaf pines. The latter two species have also been identified to be more susceptible to SPB attack than the former two (Hodges et al. 1979). If the tree resin response is not sufficient to repel the beetles and isolate the inoculum, the fungi will quickly kill the living host cells and thereby prevent further response. The fungi penetrate the ray cells and grow radially and vertically into the bole. Circumferential spread is influenced by the construction and elongation of gallery by adults and perhaps mining by larvae. Eventually a girdle of nonfunctional sapwood is effected, and the tree dies (Safranyik et al. 1975 and Coulson et al. 1980b).

Successfully attacked hosts continue photosynthesis for a period of several days to weeks after successful colonization, even though death of the host has been assured (K.W. Brown, unpublished observation). This continued photosynthesis results in a characteristic pattern of tree drying that subsequently will have an important influence on brood survival. It is not uncommon to observe an infested host, containing late-stage larvae, that still retains a green or lightly faded crown. Brood adults will later feed on fruiting bodies of the blue-stain fungus that line the walls of the pupal chamber. This maturation feeding may be a requirement for completion of development. Transfer of inoculum to the next generation is also assured.

Host Susceptibility

Although the exact mechanism controlling susceptibility to SPB attack is unknown, there seems to be little doubt that the primary host defense is the resin system. Therefore, factors that disrupt or impair functioning of the resin system will influence host susceptibility. Increased susceptibility is usually associated with reduced tree vigor, and the condition is often reflected in reduced radial growth (Coulson, Hain, and Payne 1974); phloem thickness; and resin flow rate, quality, and pressure (Helseth and Brown 1970, Hodges and Lorio 1968 and 1971, Hodges et al. 1979, and Lorio and Hodges 1968a and b).

Among the variables thought to contribute most to susceptibility are tree age, stand density, root pathogens, lightning, water imbalance, and cultural damage (Alexander 1977; Alexander, Skelly and Webb 1978; Coulson et al. 1974; Lorio 1968 and 1973; Hodges and Pickard 1971; Lorio and Hodges 1977; and Hodges and Lorio 1973 and 1975).

Variations in degree of host susceptibility influence beetle populations in several ways. First, resin flow can pitch attacking adults out. Beetles pitched out often die in the resin accumulation at the point of attack. Second, adults and eggs often die within the gallery due to resin crystallization. This phenomenon, termed resinosis, often occurs when the number of attacking adults is insufficient to successfully inoculate and kill the host. Third, when resin flow is sufficient to repel attacks, the insect is exposed to prolonged predation. Thanasimus dubius F. (Coleoptera: Cleridae), a common predator, exploits this circumstance during the spring and fall of the year. Fourth, any event that prolongs the period of time an adult is outside the host will greatly affect survival. Adult longevity between trees is short, particularly at high temperature (see fig. 5-17). Therefore, beetle confrontation with host resistance mechanisms that interfere with rapid entry into the tree usually results in insect mortality.

Habitat Suitability

The suitability of the habitat as a substrate for growth is a critical issue to beetle survival during brood development. Habitat suitability has components related to the physical characteristics of the host, such as surface area and volume of inner bark (Foltz et al. 1976b), and qualitative characteristics such as nutritive value to the developing insect (Hodges, Barras, and Mauldin 1968a and b; Hodges and Lorio 1969). These characteristics and qualities vary with tree species, age, and state of deterioration following attack. The physical properties of the host affect the rate at which the tree dries (Gaumer and Gara 1967), the temperature of the inner bark of the tree (Powell 1967), accessibility to natural enemies (Dixon and Payne 1979b), and the amount of living space available to the insect (Coulson et al. 1976b and 1980a). The chemical qualities of the habitat substrate affect the nutrients available for development of the SPB and associated arthropods, as well as the culture medium for microorganisms.

Successful colonization by SPB sets into motion a continuous process of deterioration of the habitat. An arthropod community composed of several hundred species of arthropods and microorganisms develops. This community is represented by numerous parasitoids, predators, and competitors that directly affect SPB populations. Many of these organisms have been identified (see Appendix, table 2), but most have undefined or questionable functions. Chapter 3 provides a discussion of the effects of natural enemies on populations of D. frontalis. Considerable resource partitioning occurs as a result of the development of the community. The presence of the various organisms directly influences the condition and suitability of the habitat for SPB colonization and development.

The within-tree habitat can be divided into two basic regions; the phloem (or inner bark), and the outer or corky bark. The nutrient-rich inner bark, which is utilized by virtually all members of the community, is an ephemeral habitat that changes dramatically following colonization. Conditions in the nutrient-poor outer bark are stable, relative to the inner bark, and this region is utilized sparingly.

The developmental life history of the beetle is tied closely to changes in the physical condition of the habitat. Oviposition and development of 1^st- through 3^rd-stage larvae take place in the inner bark. Development of 4^th-stage larvae, pupation, and emergence of adults usually take place in the outer bark (Goldman and Franklin 1977). This migratory behavior of 4^th-stage larvae is probably an adaptation to escape the changing conditions of the inner bark. It is not known if the migration is a response to unsuitable living conditions or unavailability of accessible food supplies. Furthermore, migration could serve as a means of avoiding natural enemies and competitors (Coulson et al. 1976b and 1979c).

Investigations of microhabitat conditions, under field conditions, have been directed primarily to studying changes in the moisture status of the tree through time (Webb and Franklin 1978 and Wagner et al. 1979). The basic pattern of SPB development in relation to measures of xylem water potential, xylem moisture, and phloem moisture is illustrated in figure 5-24. In general, development through the egg and early larval stages occurs before appreciable drying of the habitat. Migration into the outer bark occurs as drying becomes pronounced. Development of SPB life stages and the rate of deterioration of the habitat are therefore synchronized, under ideal conditions. Disruption likely results in mortality to the insect. It seems clear that habitat-related mortality is important to within-tree SPB populations, although the evidence is somewhat circumstantial. First, life table studies that include natural enemies cannot explain the amount of within-tree mortality observed. Second, attempts to rear the beetle under laboratory conditions have produced variable results. Most researchers have related the failures to problems associated with moisture condition (either too much or too little). Third, atypical elongated larval mines and failure of larvae to establish "phloem cells" have been associated with high moisture content (Webb and Franklin 1978 and Wagner et al. 1979). This condition, which may reflect unsuitable substrate conditions for growth of microorganisms (Franklin 1970b, Barras 1970 and 1973) or unfavorable conditions for insect development, results in mortality to SPB.

Figure 5-24 – Predicted phloem moisture (A), xylem moisture (B), and xylem water potential (C) at various SPB life stages, illustrating the systematic pattern of habitat degradation that follows colonization.

Many variables can influence the rate of drying of the host and the SPB’s development (Fares et al. 1980). Prominent among these variables are ambient weather conditions (Kalkstein 1976, Gagne et al. 1980a), physical characteristics of the host (Fargo et al. 1979), initial tree vigor (Lorio and Hodges 1977), and perhaps attack density. SPB adults cannot perceive relative degrees of habitat suitability at the time of colonization. Therefore, the variable age structure of the egg population likely enhances survival of the insect, as all members of the within-tree cohort are not in the same stage of development. Even though portions of the habitat might be unsuitable for development, not all life stages would be equally affected.

Host Susceptibility and Suitability in Stands

Stand dynamics and beetle population dynamics are highly interrelated. In stands, interest focuses on the distribution and abundance of the variables that were credited with influencing susceptibility and suitability in individual trees. Individual stands vary in susceptibility and suitability throughout their life. Furthermore, within a particular stand, varying degrees of susceptibility and suitability exist. Considerable research has been devoted to identifying soil, site, and stand conditions that contribute to SPB incidence (e.g., Belanger, Osgood, and Hatchell 1979b; Coulson et al. 1974; Ku, Sweeney, and Shelburne 1977; Leuschner et al. 1976; and Lorio 1968). One goal of this research has been to develop a system for risk rating stands (see Chapter 8).

The basic conclusion reached in most studies has been that SPB incidence is associated with poor tree vigor (see Hicks et al. 1978). Stand factors that contribute to poor vigor are age, density, species composition, soil texture and type, drainage patterns which lead to water imbalances, disease, and recent cultural disturbances. Within a particular stand these variables are unevenly distributed and in fact occur as a mosaic of conditions that result in varying degrees of host susceptibility and habitat suitability. Coulson (1979) provided a description of how changing conditions within a stand influence development of an SPB infestation. Microsite conditions vary substantially within a stand (Lorio 1968 and Lorio and Hodges 1971). Observed patterns of spot growth for the beetle are tied not only to number of beetles per se but also to the quantity and quality of food and habitat available to the insect. Therefore, infestation geometry, which includes both the spatial arrangement of the trees (Hedden 1978a) and susceptibility factors, is of considerable importance in spot development.

Host Susceptibility and Suitability in Forests

In the forest, we are interested in the distribution and abundance of susceptible stands. Land-use patterns and physiographic characteristics influence the general frequency of occurrence of susceptible stands. Nevertheless, the same variables that contributed to host susceptibility and suitability on individual trees and within stands operate at the forest or ecosystem level. As with individual stands, susceptible hosts within forests are characterized by clumped distributions at any one time. Forest susceptibility is a dynamic process related to the mosaic of conditions that contribute to susceptibility and suitability as well as the pattern of utilization of the forest.

Clearly, all susceptible stands within a forest do not become infested during a given SPB epidemic. Many of the variables that influence the distribution and abundance of the SPB have been described above. Availability of susceptible host type and suitable climatic conditions appear to be the basic requirements for development of beetle infestations. The SPB’s rapid appearance in response to these conditions is remarkable. Given that climatic conditions are favorable for insect development and that susceptible hosts are available in a forest, two major obstacles influence potential development of spots. First, the insect must be able to recognize susceptible host material, and second, it must be able to migrate to this material.

Primary host selection (recognition) by the beetle was discussed briefly above and in greater detail in Chapter 2. Although authorities do not agree as to the mechanism of primary host selection, two options have been proposed—directed behavior, based on visual cues or olfactory stimuli, or random searching. Obviously, olfactory stimuli would be of limited value to SPB searching large areas. Since the insect lands on host as well as nonhost species (Coster et al. 1977), it is unlikely that the beetle can discriminate between susceptible and nonsusceptible hosts by sight. This latter observation suggests that random searching is involved in initial or "coarse level" host selection.

Movement (migration) of the beetle within forests is also poorly understood. Two major hypotheses are used to explain the phenomenon: (1) migration over long distances, and (2) movement over short distances to highly susceptible hosts that serve as reservoirs. The evidence for long-distance migration is based on estimated capacity for flight, as measured by the use of flight mills, and the presence of fat reserves needed to sustain the beetles for long periods of flight (Borden 1974, Hedden and Billings 1977). But there is some evidence to suggest that long-distance migration is not the usual means of distributing populations over a forest. First, adult survival outside of the host is short (see fig. 5-17), particularly at high temperatures. Second, for a host to be successfully colonized, a large number of beetles must be aggregated in a short period of time. Third, a susceptible host apparently cannot be identified readily by the beetle. Obviously these three factors are related.

An alternative hypothesis for explaining movement of the beetle is that the insect is generally present in low numbers at many locations throughout the forest. The distribution and number of highly susceptible hosts (e.g., trees struck by lightning, infected with disease, damaged by cultural activities, injured by fire, etc.) has not been examined in detail. These trees act as reservoirs for many subcortical insects. With all the possible conditions that create highly susceptible trees, it is not difficult to envision perhaps 5 to 10 such trees occurring per 100 acres. These reservoirs would likely be able to supply limited populations from which outbreaks arise. Long-distance migration would not be a necessary requirement under such conditions.

The distribution and abundance of these reservoir trees is generally unknown, as organizations involved in survey and detection of forest pests normally do not record cause of death for single or small groups of trees. Furthermore, during periods of low population density, the distribution of SPB with the other southern pine bark beetles (I. avulsus, I. grandicollis, I. calligraphus, and D. terebrans) appears to be more equitable than during periods of high population density. Since the reservoir trees are already weakened or previously attacked by one or more of the other bark beetles, large numbers of SPB would not be required to overcome tree resistance mechanisms.

Role of SPB in Dynamics of the Pine Ecosystem

Pine forests in the South today represent second- and third-generation forests. Although the degree of management varies considerably, virtually all pine is now the result of cultivation by people. Given this circumstance, SPB has assumed (and earned) the role of a major mortality agent of pine forests. Intervention of SPB as a mortality agent in pine culture is disruptive to long-term forest management goals. Leuschner (Chapter 7) has discussed impacts of the insect in forests.

Research has documented the forest conditions associated with beetle outbreaks (see Chapter 4). Generally, the SPB is a serious pest in senescent stands occurring on poor sites. Often these stands have high basal areas and are composed of large-diameter, slow-growing trees. Of the four commercially significant pine species, loblolly and shortleaf pines are considered more vulnerable to colonization than longleaf or slash pines. In the population dynamics of commercial forests, the beetle’s major role is in killing mature pines prior to a scheduled harvest date. This mortality not only results in direct loss of revenue but also disrupts the sustained use of the forest. In addition, the beetle affects other values besides timber production, such as watershed, recreation, wildlife, and grazing (see Chapter 7).

Evaluating the beetle’s ecological role(s) in forests is somewhat more difficult than defining its social and economic impacts. Several basic problems complicate an interpretation. First, the evaluation should be directed to conditions that existed prior to the intervention of forest cultivation and management practices. Second, forests in the South are represented by more than 20 species of pine, although loblolly, shortleaf, longleaf, and slash pines are the most widely distributed species. Third, each of these tree species has characteristic ecological requirements and adaptations. Fourth, the distribution and abundance of the tree species today bears little resemblance to the pattern that existed in primitive forests. Fifth, forest management goals emphasize timber production with varying efforts to suppress mortality agents such as SPB, other insects, disease, and fire. Therefore, an evaluation of the probable roles of the beetle in forest ecosystems must be based on an interpretation of historical evidence and cast into a framework of ecological theory. Fortunately, there has been considerable research conducted in recent years on basic pattern and process of forest ecosystems (e.g., Bormann and Likens 1979).

Schowalter et al. (1979) developed an interpretative view of the role of fire and SPB in the Southeastern forest biome. This biome, which is a subdivision of the Easter deciduous forest biome, historically was prevented by fire from reaching the climax hardwood stage (Oosting 1956 and Walker 1962). Physiographically, the biome extends from coastal plains into mountain regions.

Periodic perturbation is a primary factor influencing evolution of ecosystem structure and function (Bormann and Likens 1977, Christensen and Muller 1975, Grubb 1977, Loucks 1970, and Sprugel 1976). Many ecosystems have become dependent on periodic perturbation for regeneration and cycling of limiting nutrients (Amman 1977, Christensen and Muller 1975, and Daubenmire 1974). Functionally both the SPB and fire serve as natural harvesters and as such were responsible for periodic perturbations. In combination, fire and the beetle likely maintained uneven-aged pine forests and successional openings on upland sites, as well as diversity of herbaceous, pine-hardwood, and hardwood lowland communities. These "consumers" tailor the nutrient turnover rates of the subsystems to fit resource availability and slow loss of nutrients to the marine ecosystem.

The shifting mosaic of communities within the ecosystem is important for its persistence. Ecosystem research reported by Bormann and Likens (1977) supports Loucks’ (1970) view of perturbation as means of truncating community development at a point in time prior to senescence. Senescent communities show reduced ability to regulate ecosystem function and reduced availability of r-selected or exploitive species that increase ecosystem resilience following perturbation. Fire periodically rejuvenates patches of the ecosystem by restarting development at an early stage. The SPB potentially regulates this process by (1) thinning old or stressed stands as a means of maintaining community diversity and vigor, and (2) providing concentrations of fuel to enhance the effect of subsequent fire. The resulting dynamic mosaic of communities, representing various stages of succession, increases the relative stability of the ecosystem by reducing the impact of perturbation (Bormann and Likens 1977).

The preceding scenario is based on an interpretation provided by Schowalter et al. (1980). A general view of the interaction of fire and the beetle across a gradient from lowlands to highlands is illustrated in figure 5-25. The potential roles of the SPB and fire in forest succession and nutrient cycling have not been investigated experimentally. It is likely that new insights into forest management practices could be gained by scrutiny of the historical roles of these "consumers."

Figure 5-25 – Abstraction of the Southeast coniferous forest ecosystem as a "smooth" topographical gradient (slope exaggerated). Successional transformations resulting from fire and SPB extend at right angles to the plane of the page. Dotted arrows indicate direction of movement. Fire, a regular feature of the drier uplands, invades lowlands where drought and SPB both create favorable fuel conditions. The beetle, in turn, depends on fire to regenerate pine stands. The hardwood climax reached in the far right lowland results from suitable intervals without fire and can be reduced by fire.

Role of Weather in SPB Population Dynamics

Along with the availability of susceptible and suitable host material, climate and local weather conditions in the South significantly influence the distribution and abundance of SPB. Several studies examined various aspects of the interaction of the environment and the beetle. Most have dealt with the association of weather conditions and outbreaks of the insect. The following discussion includes consideration of the effects of weather on populations occurring within trees and in spots, as well as general patterns of climatic conditions associated with outbreaks in forested areas.

Effects of Weather on Populations Within Trees and in Spots

Local weather conditions exert a significant influence on beetle survival within and between trees in a spot. Few studies actually measured weather-related mortality under field conditions. Most of the information on the subject was obtained from work oriented to describing within-tree population structure, or inferred from laboratory studies.

The most obvious effect of weather is on developmental rate of populations, which, in turn, influences the rate of spot growth. SPB population growth is continuous during much of the year, particularly in the mild, temperate regions of the Gulf Coastal Plain. This circumstance creates the basic requirement for the rapid growth of beetle spots often observed.

Both temperature and rainfall exert direct effects on survival of within-tree populations. Lethal low temperatures were reported by McClelland and Hain (1979) following severe winters in North Carolina. Likewise, Gagne et al. (1980a) reported an association of high temperature (and rainfall) with increased larval and generation mortality.

Weather exerts an indirect effect on brood survival by influencing the rate of drying of the habitat. Mortality results if the habitat dries either too rapidly or too slowly (Wagner et al. 1979). The relationship between habitat deterioration and host suitability was discussed earlier.

Adult longevity has been related to temperature conditions (Coulson et al. 1980c). Adults survive only a short time at high temperatures (fig. 5-17). Therefore, the processes of reemergence, emergence, and dispersal are greatly limited during periods of hotter weather. The effect of temperature on these processes is manifested through increased between-tree mortality and hence a reduction in the rate of spot growth.

Local weather conditions have an extremely important influence on the beetle’s ability to communicate by pheromones and attractants (see Chapter 2 for a discussion of behavioral chemicals). Fares, Sharpe, and Magnuson (1980) developed a model that demonstrates the effects of weather on dispersal of pheromones and attractants. This model explains many of the phenomena that scientists have observed concerning the aggregation patterns of the SPB in response to behavioral chemicals. For example, the pattern of diurnal activity for adults is likely a response to inversion conditions prevalent in the forest during morning and afternoon hours. Likewise the lapse conditions during midday result in funneling of the chemicals through the canopy. The first condition is extremely well suited for chemical communication: the latter is not. Furthermore, the decline in spot growth, often observed in August throughout much of the South, is likely because of the prevalence of lapse conditions during this time, coupled with high temperatures.

General Patterns of Climate and SPB Outbreaks

Indices of temperature and rainfall and wind direction have been compared to relative estimates of beetle populations in order to identify conditions that influence growth and collapse of outbreaks (Wyman 1924, Craighead 1925, Beal 1927 and 1933, and St. George 1930, Merkel 1956, King 1972, and Kalkstein 1976). These studies have not provided a single set of conditions that are consistently associated with outbreaks in different sections of the South or at different times in the same section. The type of data used in the studies has contributed to the inconsistent results obtained.

Surveys of SPB-caused damage have provided most of the data used to characterize beetle populations. Survey data may be inaccurate for this specific application because crowns of infested pines fade at different rates (Doggett 1971, Billings and Kibbe 1978), there are large and generally undefined observation errors in collecting the data (Mayyasi et al. 1975), small infestations (< 10 trees) are often not reported by surveying agencies, and there is no simple and direct relationship between the number of dead or faded trees and beetle populations (Thatcher and Pickard 1964 and 1967).

The weather data used in the studies, which generally consisted of temperature and rainfall information, were more uniform in quality than insect population data, because the former were taken at weather stations and standard procedures were used to calculate indices. However, these indices were often expressed as deviations from regional averages. King (1972) pointed out that beetle-host interactions may not be governed by simple deviations from average conditions.

The temperature data from the weather stations probably provided a reasonable approximation of conditions in nearby forests, given that topography was similar. But a forest is generally 1 to 3 degrees C cooler than an open field during the day and 1 to 3 degrees C warmer at night (Geiger 1957). Rainfall is more variable locally than temperature. The correspondence between rainfall beneath the forest canopy and in an open field is further complicated by stemflow and interception of rain by tree crowns. Therefore, measurements of rainfall in a local area probably do not correspond well with measurements taken at weather stations several kilometers away.

The final point regarding the type of data used is that the weather variables were likely too simple to capture the complex relationships of the SPB and its environment. As indicated earlier, the numerical expression of beetle populations at the ecosystem level is the result of complex interactions among variables. So it is not surprising that studies of climatic patterns have not revealed consistent results suitable for explaining the SPB’s distribution and abundance. It should be emphasized also that most of the studies on weather and SPB outbreaks have not been directed to explaining the effects of weather conditions on the insect per se. Rather, weather variable were interpreted to influence host susceptibility and suitability, which, in turn, were related to the occurrence of outbreaks.

Two studies have utilized absolute estimates of within-tree beetle population in conjunction with weather station data on temperature (McClelland and Hain 1979) and temperature and rainfall (Gagne et al. 1980a). In the first study, conducted over a 2-year period in North Carolina, severe winter temperatures resulted in nearly 100 percent brood mortality and the subsequent collapse of spot growth. During 1 year of the study, however, winter temperatures were not low enough to kill brood life stages, and yet the infestations still collapsed.

In the Gagne study, survival of larvae and generation survival were both associated with temperature and rainfall indices over a 3-year period in east Texas. The most useful indices in regression equations describing larval survival and generation survival were day-degree accumulation and the proportion of the infestation period when at least 0.6 cm or more of rain fell per day. Increases in both of these indices were associated with decreases in larval and generation survival. The authors believe these indices the most useful for describing associations between weather and beetle populations because they measure both the size and duration of temperature or rainfall. The interaction of the two indices was complementary (Gagne et al. 1980a). This study did not substantiate King’s findings (1972), which suggested that outbreaks were preceded by above-normal rainfall in January and February, or the findings of Kalkstein (1976), which suggested that outbreaks were triggered by dry soil conditions.

Implications of SPB Population Dynamics in Pest Management Decisionmaking

Integrated pest management (IPM) has been defined in a number of ways (see Chapter 11). Simply, IPM is the maintenance of destructive agents, including insects, at tolerable levels by the planned use of a variety of preventive, suppressive, or regulatory techniques and strategies that are ecologically and economically efficient (Waters 1974). IPM, a component of total resource planning, has evolved during the last decade (Coster 1977). Generally an IPM system for insects in a forest has four basic components that must be defined and understood: (1) pest population dynamics, (2) forest stand dynamics, (3) treatment tactics and strategies, and (4) impacts. Each of these subjects is discussed in various chapters in this volume.

Information on Population Dynamics in Relation to Decionmaking in Pest Management

In this chapter, the SPB’s population system was presented in a hierarchy of increasing complexity beginning with events taking place at or in individual trees, progressing to spots (or infestations), and ending at the ecosystem level. The spot level of complexity was highlighted because it is at this point where actual dynamic features of both the insect and host systems come into play. Much of the accelerated research on SPB-host dynamics has been undertaken because we now recognize the vital roles of these components in pest management decisionmaking. The problem of beetle-induced tree mortality in forests is a function of the distribution and abundance of both trees and beetles through space and time. The mosaic patterns of susceptible and suitable hosts in forests and the many variables that influence beetle population numbers have been described above. Given this complexity, one can appreciate the difficult problem faced by forest managers in predicting when and where spots will occur and whether they will increase in size or become inactive. Furthermore, evaluating the efficacy of suppression tactics and prevention techniques is not a simple matter: it requires sophisticated understanding of the interaction of the host-insect systems (Coulson et al. 1979c).

Because of the extreme complexity of the beetle-host systems, researchers have developed sophisticated mathematical models to predict populations of the SPB and the timber mortality resulting from its activity (see Chapter 6). The models can also be used in evaluating efficacy of treatment tactics aimed at suppression of populations. Regarding population dynamics, there is a distinct relationship between understanding, prediction, and decisionmaking (Campbell 1973, Coulson 1974). Mathematical models, based on detailed understanding of the population system of the SPB and interaction of this system on the host, enhance decisionmaking capability.

Population Dynamics Information and Evaluation of Treatment Tactics

If we know how the population system of the SPB operates, we can develop and test new treatment tactics proposed for the insect. Scrutiny of information on the beetle’s population dynamics has revealed a highly evolved and complex array of survival mechanisms, including (1) density-dependent regulation of egg populations, (2) reemergence of parent adults, (3) blending of emerged brood adults and reemerged parent adults to form the attacking adult population, (4) incremental allocation of both emerged and reemerged adults, (5) migratory behavior of within-tree larval populations possibly to escape competitors and changing habitat conditions, (6) variable age distribution of within-tree life stages, (7) coutilization of hosts by several species of bark beetles during periods of low population numbers, (8) utilization of highly susceptible hosts, and (9) communication via behavioral chemicals. There are probably many more mechanisms for SPB survival that have not been identified. The important point is that since the SPB is a native pest that coevolved with its host, it should be expected that elaborate survival mechanisms exist which enhance perpetuation of both the insect and the host species.

If a goal of forest management is to maintain pest populations at tolerable levels through application of remedial tactics, then it is imperative that the survival mechanisms of the insect be considered. A treatment tactic can be viewed simply as another mortality agent imposed on the SPB life system. Knowledge of how the beetle responds naturally to other mortality agents provides insight into the probable success or failure of a proposed treatment tactic. Mathematical models of SPB population dynamics are the tools for such evaluations. Failure of past suppression projects against SPB can be directly attributed, in many cases, to lack of understanding of the survival mechanisms of the insect and to our inability to predict the outcome of a treatment on the population system.

Acknowledgements

I wish to acknowledge and thank my colleagues at Texas A. & M. University (P.E. Pulley, T.D. Schowalter, T.L. Wagner, J.A. Gagne, W.S. Fargo, D.N. Pope, J.D. Cover, A.M. Bunting, P.B. Hennier, T.L. Payne, G.L. Curry, P.J.H. Sharpe, R.M. Feldman, Y. Fares, and K.W. Brown) for their aid in the formulation of this review and for permission to use unpublished data. Special thanks are extended to members of the "Population Dynamics Working Group" of the ESPBRAP, who were responsible for advancing our understanding of population dynamics of D. frontalis. In particular I am indebted to F.P. Hain (North Carolina State University), F.M. Stephen (University of Arkansas), T.E. Nebeker (Mississippi State University), C.J. DeMars (U.S. Forest Service), and J.E. Coster (U.S. Forest Service).

Some of the findings reported herein were supported by the ESPBRAP projects TEX06163-G and TEX06355; the Texas Agricultural Experiment Station, TEX6009; and the National Science Foundation, NSFGB34718.

Special thanks are extended to T.E. Nebeker for providing criticism of an earlier draft of this manuscript, to P.B. Hennier for preparation of the figures used herein, to A.M. Bunting for editorial assistance, and to Connie Fisher for clerical assistance.

This paper is contribution TA 15684 from the Texas Agricultural Experiment Station.

[  Contents  ]   [  Previous  ]   [  Next  ]   [  Home  ]