Paul E. Pulley

Heterogeneity of forest landscapes and the distribution and abundance of the southern pine beetle.

A fundamental and unanswered research question in landscape ecology centers on how the spatial arrangement of ecosystems influences the distribution and abundance of organisms across complex landscape mosaics. In this study our goal was to examine how the southern pine beetle (SPB) perceives and responds to heterogeneity in forest landscape mosaics. The study was based on the use of extant knowledge of SPBs natural history and a spatially referenced database that included explicit information on landscape structure as well as distribution and abundance of the insect. Both the content and context of the spatial elements forming the forest landscape were considered, as human-caused fragmentation and natural disturbances create mosaic patterns where the specific arrangement of components can enhance or inhibit herbivory by the SPB. To examine how heterogeneity influences epidemiology of the SPB, we identified the elements of landscape structure that serve as targets for SPB. The principal targets include three types or arrangements of host trees: acceptable species, susceptible habitat patches, and lightning-struck hosts. Using a spatial database and a statistical approach we evaluated the number and arrangement of these targets in the context of existing population centers. The product was a functional heterogeneity map that portrayed how the spatial arrangement of landscape elements (habitat targets) influences the distribution and abundance of the SPB across complex mosaics of ecosystems.

Impact of foraging byMonochamus titillator [Col.: Cerambycidae] on within-tree populations ofDendroctonus frontalis [Col.: Scolytidae]

The impact of foraging byMonochamus titillator (Fab.) on within-tree populations ofDendroctonus frontalisZimm. was described. Total population estimates forD. frontalis and inner bark area foraged byM. titillator were computed for 89 loblolly pine,Pinus taeda L., sampled over a 3 year period. Frequency histograms were prepared for the surface area of habitat infested, area foraged byM. titillator, and within-tree populations ofD. frontalis. These data were used to calculate the proportion of area foraged byM. titillator, the proportionalD. frontalis mortality for the entire tree, and the proportional mortality occurring in the foraged area. Histograms showing the variation of these components were prepared and described using nonlinear mathematical models. Ca. 20% of the infested surface area was foraged byM. titillator. Mortality toD. frontalis on a per tree basis was ca. 14%. Mortality in the area foraged byM. titillator was ca. 70%. These estimates were highly variable between individual trees and a procedure for predicting the probability of a given level of foraging and mortality was described. Sources of variation influencing foraging byM. titillator and mortality toD. frontalis were investigated. Variation between years and season followed similar trends with both foraging and mortality increasing from the base to the top of the infested bole. Variation between tree size-class was highly significant. Larger size-class trees had substantially greater foraging and mortality than did the smaller size-classes. Mortality within the foraged area was also found to be greater at the extremes of the infested bole.RésuméL’impact de l’activité deMonochamus titillator (Fab.) sur les populations à l’intérieur de l’arbre deDendroctonus frontalisZimm. est décrit. Le total des évaluations de population deD. frontalis dans des zones à l’intérieur du tronc attaquées parM. titillator a été soumis à l’ordinateur pour 89 pins,Pinus taeda L., échantillonnés pendant 3 ans. Les histogrammes de fréquence ont été établis pour la surface de l’habitat infesté, la surface explorée parM. titillator et pour les populations deD. frontalis présentes dans l’arbre. A partir de ces données on a calculé la part de surface utilisée parM. titillator, la mortalité deD. frontalis extrapolée pour l’arbre entier et la mortalité se produisant dans la surface exploitée. Les histogrammes traduisant les variations de ces composantes ont été dressés en ayant recours à des modèles mathématiques non linéaires. Environ 20% de la surface infestée a été exploitée parM. titillator; la mortalité du scolyte fut d’environ 14% dans l’arbre et de 70% dans la zone attaquée parM. titillator. Ces évaluations sont très variables d’un arbre à l’autre et on a mis au point un procédé pour prévoir la probabilité d’un taux donné d’activité alimentaire du cerambycide et de mortalité du scolyte. Les causes de cette variabilité ont été étudiées. La variation suivant les années et les saisons suit les mêmes tendances à savoir l’augmentation à la fois de l’activité alimentaire et de la mortalité de la base au sommet du pin attaqué. La variation selon les classes de taille des arbres est hautement significative: les arbres les plus gros présentent une exploitation et une mortalité beaucoup plus importantes que les arbres plus petits. La mortalité du scolyte dans la zone attaquée par le cerambycide est également plus élevée aux extrémités du pin infesté.

Continuous population estimates forDendroctonus frontalis Zimm. (Coleoptera: Scolytidae) occurring in infestations

Infestations ofDendroctonus frontalis Zimm. are often observed to enlarge continuously by the colonization of new hosts in a pattern similar to a forest fire. This pattern of infestation growth presents unique problems in quantitatively estimating populations ofD. frontalis. Beetle populations on each infested tree in an infestation go through five processes: attack, oviposition, reemergence, survivorship, and emergence. These processes, which have been described mathematically in the literature, each take several days for completion. In order to follow the distribution and abundance ofD. frontalis throughout the course of development of a spot, we need a daily estimate of the number of beetles involved in each process on every tree. Since it is not practical to sample each tree daily, we developed a procedure whereby quantitative estimation procedures for within-tree populations were used in combination with the mathematical models for the life processes to produce a daily record of the number of adults successfully attacking trees, the number of eggs oviposited, the number of beetles reemerging, number of beetles surviving within the trees, and the number of beetles emerging. These daily estimates were then summarized for all trees in the spot for the duration of the infestation. The daily record of populations ofD. frontalis, used with information on infestation geometry, were suggested to be of value in describing and elucidating several important facets of population dynamics including dispersal patterns within infestations, between tree beetle loss (mortality), and time lags among the various population processes. The information reported can be used to develop simulation models of population dynamics or to validate existing models.

Mathematical description of within-tree distributions of the various developmental stages ofDendroctonus frontalis Zimm. (Coleoptera: Scolytidae)

Probability density functions for the various within-tree developmental life stages ofD. frontalis are derived. The differences between some of these distributions are discussed. The possibility of the existence of a negative feed-back mechanism which enables the beetle to make optimum utilization of the available resources is proposed. In addition, the concept of an optimum location on the infested bole where the environmental (tree) conditions are such that maximum survivorship occurs is advanced.

SAMPLING CONSIDERATIONS FOR EVALUATING THE EFFECTS OF MORTALITY AGENTS ON BARK BEETLES

Basic requirements for estimating effects of mortality agents on bark beetle populations are examined. Emphasis is placed on sampling within-tree populations of both host insect and mortality agent at single points in time. The estimation procedures described should be suitable for obtaining information on the effects of mortality agents on the distribution and abundance of the host insect through space and time. When multiple samples are to be collected the topological estimation procedure described is superior to the other options presented. The topological procedure is suitable for obtaining estimates of surface area or volume of the habitat, density of the host insect, and density of the mortality agent. An example of interspecific competition between Dendroctonus frontalis and Monochamus titillator is examined. This example is comparable to woodpecker predation of D. frontalis from both logistical and analytical sampling standpoints.

Accuracy and precision of the topological mapping procedure for estimating within-tree populations of bark beetles

The accuracy and precision of the topological mapping procedure for estimating within-tree populations of bark beetles was investigated for a variety of different sampling conditions. Simulation techniques were used to define the mean and standard deviation of proportional errors encountered in estimation under different sampling intensities. The number of samples collected at a particular height and the vertical spacing between heights were varied. Information presented should aid in developing sampling plans for studies of bark beetle populations and will permit the reexamination and/or recovery of historical data sets on bark beetle populations.

Estimatingdendroctonus frontalis (Coleoptera: Scolytidae) daily infestation dynamics

Mathematical procedures are given to estimate infestation totals and daily life stage arrivals, departures, and mortality ofDendroctonus frontalis Zimmermann for an infested tree in the field. These estimates are based on minimal sample data and are designed to utilize all available information. Daily arrival estimates for larvae, pupae, and callow adults are obtained by indirect analysis without direct observation of these stages. The procedures are applied to 147 infested trees, and the results are transformed to a common time basis to obtain daily expectations by life stage for an “average” tree. These expectations suggest optimal times for field sampling or relative times of sampling when optimal times are missed. Expected daily arrival distributions by life stage for a single egg and a single attacking adult are given. Procedures are given for utilizing collateral information to obtain an infestation total and daily arrival estimates for a boundary life stage. The results of this study are applicable to anyD. frontalis field study, and the procedures given are applicable to any bark inhabiting insect having similar habits.

Calculation of quasi-hydrostatic restoration coefficients for complex configurations

This paper describes general capability and typical uses of a new computer system developed for calculation of the non-linear quasi-hydrostatic reactions of complex hulls, columns, braces, etc. typically used in the design of semisubmersibles, pipe-lay stingers and other marine configurations.