No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Defoliation
Abstract
From ancient times, mankind has attempted either to harness or to prevent defoliation of trees and plants. The positive and negative impact of defoliation by invertebrates (chiefly insects), vertebrates, plant diseases, or abiotic factors such as air pollution is an integral part of agriculture, forestry, and natural resource management. Defoliation is “an unseasonable reduction in the foliage cover of a plant due to attacks by insects or fungal disease, or as a result of other factors such as drought, storms, or chemicals in the atmosphere.” Defoliation encompasses partial or complete loss of the plant's foliage and individual leaves. The most significant impacts of natural defoliation are crop losses in agriculture and production forestry and, for native plant communities, possible changes in structure and species composition and loss of amenity values. It is not possible to review in this article the huge amount of information available on defoliation. In terms of measurement, estimation, or prediction, there are a number of commonalties across the range of activities in which quantification of defoliation is required.
ResearchGate has not been able to resolve any citations for this publication.
This paper explores the use of economic thresholds as an aid to decision making in forest pest management. Economic thresholds have been applied to insect, weed and fungus control problems in agriculture. The motivation for the use of economic thresholds in pest control is both the reduction of the total amount of pest control product used and the more cost-effective use of the product that is applied. To date, however, the application of economic thresholds has been limited in forestry. Extensions of the available framework, particularly extensions to accommodate the temporal dimensions of forest production, are described. Issues related to data needs are identified.
Changes in forage utilization have been difficult to measure non-destructively without some level of subjectivity. This subjectivity, combined with a lack of reproducibility of visual estimates, has made forage utilization measurement techniques a topic of considerable discussion. The objective of this study was to develop and test the accuracy and repeatability of an objective, computer-based technique for measuring changes in plant biomass. Digital photographs of target plants acquired before and after partial defoliation were analyzed using readily available image analysis software. Resulting data were used to develop a simple linear random coefficient model (RC) for estimation of plant biomass removed based on the area of the plant in the photo. Sample collection took approximately 20 minutes/plant for alfalfa (Medicago sativa L.). Analysis of images took another 60 to 90 minutes. Regression analysis gave an R-2 of 0.969 for predicted vs. observed plant weights. Testing this model using 10 alfalfa plants yielded weight estimates of defoliated plants accurate to within +/- 8.5%. The advantage of the RC model is its ability to use easily obtained coefficients from simple linear regression models developed from each plant in a way that accounts for the lack of independence between samples within an individual plant. The technique described here offers an objective and accurate method for measuring changes in plant biomass with possible applications in ecology, botany, and range science. In particular, application of this technique for estimating forage utilization may improve accuracy of estimates and, thereby, improve range management practices.
The influence of defoliation type (artificial versus natural), timing (early versus late), and intensity (25%, 50%, and 75% of needle mass removed) on leader growth of Scots pine, Pinus sylvestris (Linnaeus), was assessed for 2 years after treatment on an even-aged stand located in southeastern Finland. Trees were defoliated simultaneously, either artificially with a pair of scissors or naturally with larvae of Neodiprion sertifer (Geoffroy) and Diprion pini (Linnaeus) for the early- and late-season treatments, respectively. After 1 year, early-season artificial defoliation generally caused greater growth reduction than natural defoliation. Late-season defoliation yielded opposite results. Trees defoliated artificially in early-season treatments were significantly shorter than control trees irrespective of defoliation intensity, whereas those defoliated late in the season did not differ from controls, except at the highest intensity. Trees defoliated by sawflies, either early or late in the season, were significantly shorter than control trees only at the highest defoliation intensity. The pattern of growth loss in the second year appeared similar to that in the first year. The impact of defoliation was either prolonged neutral or negative, as no compensatory responses on height growth in Scots pine were observed. Timing of the treatment in relation to completion of leader growth, differences in defoliation types, alteration of the photosynthetic capacity due to biomass loss, and the functional role of plant parts defoliated may explain the results observed.
Quantitative assessments of the incidence and severity of Mycosphaerella leaf disease were made on nine provenances (encompassing the four subspecies) of Eucalyptus globulus Labill. over three seasons in 1990 in a trial in Victoria, Australia. Defoliation was also assessed and tree height and diameter measured. Mycosphaerella leaf disease increased rapidly from winter to spring with disease incidence reaching 100% in most provenances by summer. There were highly significant differences in disease severity among provenances, with provenances from E. globulus ssp. globulus Kirkpatrick and E. globulus ssp. bicostata (Maid. et al.) Kirkpatrick being the most severely affected, while provenances from E. globulus ssp. maidenii (F. Muell.) Kirkpatrick and E. globulus ssp. pseudoglobulus Kirkpatrick were only slightly affected. There were also highly significant differences among provenances within E. globulus ssp. globulus and E. globulus ssp. pseudoglobulus. There was a high correlation between disease severity in summer and defoliation, suggesting that Mycosphaerella leaf disease was the main causal agent in leaf fall. There was also a high negative correlation between disease severity and growth rate. This study shows the potential benefits of selecting resistant provenances of E. globulus to maximise production from plantations in areas where Mycosphaerella leaf disease is a problem.
The approaches for modeling yields of crops in relation to plant disease intensity are reviewed. It is shown how these models can be expanded or modified to account for: i) duration of healthy leaf area during a growing season and the interception of radiation by leaf area; ii) spatial pattern of pathogens or diseases and resulting mean yield; and iii) time of infection of plants (especially by viruses and other organisms that cause systemic infection) linked to the temporal dynamics of disease incidence in fields. These expansions or modifications can make predictions of crop losses more precise or increase our understanding of the disease/loss relationship. Areas where more research is needed are discussed, especially: i) dealing with error in measuring disease intensity; ii) validating models; and iii) relating yield to multiple diseases or pests.
Plants in natural systems that can tolerate or compensate (recover equivalent yield or fitness) for herbivore feeding have selective advantages that lead to genotypic maintenance. In agriculture, too, these attributes are of value, especially to yield. Key problems in identifying plant comepnsation are outlined: it is stressed that there is a lack of a consistent linear relationship between carbohydrate production and growth/yield, a consequence of the complexity and variability of the plant resource-allocation infrastructure. Endogenous factors affecting plant compensation include regrowth patterns, photosynthetic activity, senescence, leaf morphology and canopy architecture. Exogenous factors include nutrient availability, intensity and timing of defoliation, herbivore distribution, and water and temperature stress. -P.J.Jarvis
Leaf area index (LAI) after insect defoliation injury is closely related to yield, so measuring the LAI of defoliated plant canopies is of considerable importance. Unfortunately, direct measurement of LAI is laborious and time-consuming. This study was conducted to evaluate the LAI-2000, a commercially available instrument for rapidly and accurately measuring LAI of insect-defoliated soybeans, Glycine max (L.) Merr. The instrument's indirect measurements of LAI were compared with direct measurement of LAI. Measurements were made in 2 fields, one with ≈50% defoliation and the other with ≈30% defoliation. The LAI range for direct measurement was 2.47-4.24. The instrument's indirect LAI measurements exceeded the direct measurements, but by <15%. With respect to plant response to insect defoliation, an error rate of ≈15% is acceptable for most applied and research purposes. Additional comparison of LAI-2000 estimates to actual LAIs from 3 experiments revealed a linear relationship with a slope not significantly different from 1. Therefore, our work indicates the LAI-2000 provides accurate estimates of LAIs for insect defoliated soybeans, and would likely perform comparably in other crops.
L'exactitude et la precision des notations de maladies peuvent etre ameliorees en selectionnant les methodes les plus appropriees et en donnant la formation necessaire aux evaluateurs. Jusqu'ici, tres peu d'etudes ont ete faites pour aider les chercheurs a choisir la methode ou le systeme de notation des maladies qui va le mieux satisfaire leurs exigences en terme d'exactitude, de precision et des ressources disponibles. Le choix de la methode d'evaluation est tres important, considerant le cout eleve de la recherche sur le terrain. L'utilisation de modeles et methodes statistiques tres perfectionnes pour traiter des notations de variation inconnue a aussi ete mise en doute. En plus d'utiliser l'analyse de variance, l'analyse de correlation et le coefficient de variation, pour evaluer et comparer les methodes de notation et les evaluateurs, la regression du moindre carre peut etre utilisee pour determiner s'il y a une relation lineaire significative entre des notations repetees par le meme evaluateur (reproduction intra-evaluateurs) et s'il y a une relation significative entre les notations par differents evaluateurs (correlation entre-evaluateurs). Les parametres de l'equation des regressions (pente, point d'intersection, erreur type de l'estime-Y, coefficient de determination) peuvent etre utilises pour comparer l'exactitude et la precision des methodes et des evaluateurs. L'exactitude des evaluations peut etre amelioree avec de l'entrainement approprie. Le logiciel Disease. Pro, programme de formation pour evaluer les maladies, a ete developpe pour aider les evaluateurs a ameliorer leur capacite a estimer l'intensite des maladies. Apres une seance d'une heure de formation avec Disease. Pro, 70 evaluateurs sur 80 ont ameliore significativement leur habilite a estimer l'intensite de la tache tardive sur les feuilles d'arachide.
Since Noy-Meir’s (1975) analysis, much progress has been made in refining understanding of the grazing process in temperate grassland. Particularly, advances have been made in understanding grazing at the bite scale (Illius & Gordon 1987; Spalinger & Hobbs 1992; Parsons et al. 1994). Domestic ruminants remove a relatively fixed proportion (some 40–70%) of standing vegetation with each bite (Ungar, Genizi & Demment 1991; Laca et al. 1992). If ruminants harvested 50% of the standing vegetation uniformly across an entire grazeable area each day, this would exceed the ability of the vegetation to recover from defoliation (the estimated sustainable homogeneous removal rate is 2·5% of standing vegetation per day; Parsons & Chapman 1998). Thus, for a grazing system to be sustainable, grazing must take place from only a small portion of the total grazeable area each day (some 5%; Wade 1991; Parsons & Chapman 1998). This simple calculation illustrates that grazed grassland is necessarily heterogeneous, because at any time, different patches will be in different states of recovery from previous defoliations. As has long been recognized, ‘continuous’ grazing is not a continuous process, but involves a succession of discrete defoliations at the bite scale, each followed by a variable period of regrowth (Hodgson & Ollerenshaw 1969; Morris 1969; Clark et al. 1984). Furthermore, it is not a deterministic process: because animals take bites from only a small part of the total area on any day, there is some uncertainty as to exactly where animals choose to take bites. Thus, when considering grazing at the bite scale, we must deviate from Noy-Meir’s assumptions in three ways: we must assume (1) spatial heterogeneity instead of homogeneity, (2) discrete instead of continuous defoliation and (3) stochasticity instead of determinism.
Disturbances such as fire, hurricanes, and herbivory often result in the net release of CO2 from forests to the atmosphere, but the magnitude of carbon (C) loss is poorly quantified and difficult to predict. Here, we investigate the carbon balance of an oak/pine forest in the New Jersey Pine Barrens using the Canopy Conductance Constrained Carbon Assimilation (4C-A) model. The 4C-A model utilizes whole-tree sap-flux and leaf-level photosynthetic gas exchange measurements at distinct canopy levels to estimate canopy assimilation. After model parameterization, sensitivity analyses, and evaluation against eddy flux measurements made in 2006, the model was used to predict C assimilation for an undisturbed year in 2005, and in 2007 when the stand was completely defoliated for 2–3 weeks during an infestation of gypsy moths (Lymantria dispar L.). Following defoliation, only 50% of the foliage reemerged in a second flush. In 2007, canopy net assimilation (AnC), as modeled with the 4C-A, was reduced to approximately 75% of AnC in 2006 (940 vs. 1240 g C m−2 a−1). Overall, net primary production (NPP) in 2007 was approximately 240 g C m−2 a−1 (vs. 250 g C m−2 a−1 in 2006), with 60% of NPP allocated to foliage production, a short-term carbon pool. Woody biomass accumulation, a long-term carbon pool, was reduced by 20% compared with the previous year (72 vs. 57 g C m−2 a−1 in 2006 and 2007, respectively). The overall impact of the defoliation spanned 21% of upland forests (320 km2) in the New Jersey Pine Barrens, representing a significant amount of overall C not being taken up from the atmosphere by the forest, thus not accumulated in the biosphere.
Defoliation of Eucalyptus regnans plantations in Tasmania by the leaf beetle Chrysophtharta bimaculata (Olivier) over a two year period significantly reduced growth of one-year-old previously uninfested trees and six-year-old trees with a history of defoliation.One-year-old trees defoliated by C. bimaculata had lost 45.6% and 52.1% of their potential height increment and basal area increment respectively by the end of the two year trial compared to protected trees. Six-year-old trees experienced similar growth losses and the biomass of leaves and stems/branches of unprotected trees at the end of the two year period was reduced by 44% and 39% respectively compared to protected trees.The impact of the defoliation levels recorded in these trials is discussed in relation to the current control strategy for C. bimaculata and the economics of eucalypt plantations in Tasmania.
A mathematical model has been developed for estimating the consumption of canopy vegetation by insects. The model divides canopy leaves into age classes according to time of leaf emergence. Periodic measurements were made of gross leaf area and "hole" area initiated by insect consumption for each generation. Gross area distributions were used to determine empirical growth functions for the canopy vegetation. The rate of change in hole area is a function of both insect consumption and subsequent hole expansion due to leaf growth. The hole expansion rate was determined empirically by punching holes in known area in growing leaves and correlating hole expansion with leaf growth. From these relationships, a linear differential equation was derived to describe the dynamics of insect grazing in the canopy. This model was used to interpret the effect of insect herbivory on the dry matter and nutrient cycles in a Liriodendron tulipifera forest. Average herbivorous insect consumption over 3 years was equivalent to 2.6% of the net primary production of foliage biomass but represented an annual loss of 7.7% in photosynthetic surface area.
Facilitated by advances in analytical chemistry, soil, plant and water analysis now play a key role in the qualitative and quantitative characterisation of the environment. Ecological and health risk assessments can predict probabilities and risks of nominated environmental outcomes from the analytical data. The accuracy of these predictions, however, is dependent on data interpretability, which is constrained by our ability to determine that fraction of the contaminant that is active in biological systems. Soil analyses, for example, are used to characterise chemically contaminated sites and determine remediation thresholds. Such decisions, though, typically are based on total analyte concentrations that often are unrelated to environmental risk. While combined ecotoxicity and chemistry tests enable quantification of soil toxicity and identification of the analytes responsible, further progress may depend on development of sequential extraction and speciation tests calibrated against bioavailability/toxicity. Similarly, while we can measure most contaminants in plant tissues and understand how they affect plant metabolism, we remain unable to identify/quantify the physiologically active fraction. There is scope to improve interpretation by linking sub‐cellular chemical forms to toxicity. Substantial gains, too, have been made in water testing with the development of ambient water quality criteria related to organism‐level responses. We have much to learn, though, about effects of speciation, complexation, sediment/water column interactions, and ecosystem‐level responses. Solving such challenges should expand opportunities for environmental managers to use soil, plant and water analysis in the 21st Century.
The optimum number of leaves and terminals to sample to estimate the incidence and severity of Alternaria blotch, caused by Alternaria mali, of apple (Malus × domestica) and associated defoliation was derived from data collected between 1989 and 1992 in three locations in North Carolina. On the basis of estimates of variance and cost, the mean optimum number of leaves per terminal was 16.88 for evaluating disease severity and 19.37 for determining disease incidence. The optimum numbers of terminals per tree for assessing severity, incidence, and defoliation were 1.2, 1.18, and 2.04, respectively. The variation in disease among leaves and among terminals contributed equally to overall estimates of variance [...]
Threshold models can be useful for analysing ordered categorical data, like ratings. Such models provide a link between the ordinal scale of measurement and a linear scale on which treatments are supposed to act. In this paper a simple agricultural plot experiment is considered with two sources of variation, namely between-plot variation and within-plot variation. So far, methods for analysing ordered categorical data are not capable of handling such a situation adequately. It is shown that for a threshold model with two sources of variation maximum likelihood estimates can be obtained by iterative weighted least squares. The computer package Genstat is used to carry out the computations. To illustrate the methods an example concerning damage in strawberries due to the fungus Phytophtora fragariae is given.
On-site climatic measurements were used to model red pine (Pinus resinosa Ait.) shoot elongation. Three study sites each consisting of three 0.2-ha plots were cleared and planted with red pine. Shoot growth was measured weekly for 2 years. Incremental seasonal growth of the leading shoot was estimated using a difference form of a modified Chapman-Richards growth function. Weekly growth was estimated as a function of air temperature degree days (4.4°C basis), soil water potential, and total expected seasonal growth. An example using the model compares varying site and climatic conditions and their effect on the pattern of seedling height growth during the growing season as well as their effect on the total amount of height growth realized at the end of the growing season.
Tropical cyclones (“typhoons”) affect ecosystem processes by disturbing the ecosystem's biogeochemistry, population dynamics, and services. The impact of 10 typhoons that struck the Japanese Islands in 2004 was studied using a process-based ecosystem model (Vegetation Integrative SImulator for Trace gases) and by interpreting deviations in the carbon budget observed at the Takayama forest site in central Japan. The site-calibrated model appropriately simulated gross and net CO2 fluxes in most years but could not capture the clear depression of CO2 uptake in 2004, probably because it neglected the impact of defoliation on the canopy's carbon gain due to strong typhoon winds. The defoliation intensity caused by each typhoon event was inversely estimated from the flux measurement data and using a Monte Carlo approach. Accounting for the repeated 10%–20% defoliation that occurred in 2004 lowered the canopy carbon gain by nearly 200 g C m−2 yr−1, resulting in better agreement between the estimated and observed flux values. Comparison of the estimated defoliation pattern with satellite-based estimates of leaf area index indicated that such moderate defoliation from midsummer to autumn was plausible. The influence of CO2 emissions from typhoon-generated debris on the carbon budget in subsequent years was estimated to be tiny in this case. These results have implications for regional carbon accounting studies, because both devastating and moderate defoliation caused by tropical cyclones can have a marked effect on the regional carbon budget, especially when defoliation occurs repeatedly within a single growing season.
Feeding rates of Colorado potato beetle larvae were measured in 24-h trials in small field cases on potato leaflets that remained attached to the plants. Four models were constructed to explain the results; these differed by including or excluding the effect of insolative heating on larval body temperature, and by including or excluding thermoregulatory behavior, in all combinations. In all instars, observed feeding was independent of mean body temperature during the trial as estimated by each model, and less than the amount predicted using this mean to drive constant-temperature feeding rate functions.
The four models were assessed by regressing model predictions on observed feeding rates. For first and second instars, there was no relationship between observed feeding and the amount predicted by any of the models, probably because the feeding rate was small relative to the precision of the measurements. For third and fourth instars, predictions from the model that included both insolative heating and behavioral thermoregulation did not differ significantly from observations. Simulations using this model suggested that daily mean body temperature fails to predict foliage consumption because this mean cannot account for the effects of transient occurrences of suboptimal temperatures.
Recurrent restricted phenotypic selection (RRPS) has increased spaced-plant yield of 'Pensacola' bahiagrass (Paspalum notatum Flugge var. saurae Parodi) and led to higher-yielding, more erect genotypes. Response to defoliation was examined in three entries: selection Cycle 0, Pensacola; Cycle 9, 'Tifton 9', and Cycle 14. In a 3-yr plot study, two cutting heights (15 or 100 mm) in combination with three regrowth intervals (2, 4, or 8 wk) were imposed throughout 24 wk. A split-split plot arrangement of regrowth interval (whole plot), cutting height (subplot) and entry (sub-subplot) was used, with six replications. In a 2-yr grazing study, response of the same entries to defoliation from continuous stocking was evaluated. In the plot study, yield (7030 kg ha-1) was maximized by low cutting height and 8-wk regrowth interval during the first year. In year 2, a 4-wk regrowth interval and low cutting height produced the highest yield (11 220 kg ha-1). Mean yields of Tifton 9 (7650 kg ha-1) and RRPS cycle 14 (7320 kg ha-1) were greater than Pensacola (6500 kg ha-1), but response of entries to cutting height and regrowth interval varied among years. In the grazing study, herbage mass was greater for Tifton 9 (1670 kg ha-1) and RRPS Cycle 14 (2000 kg ha-1) than for Pensacola (1420 kg ha-1). Carrying capacity of Tifton 9 (935 d ha-1) was greater than Pensacola (855 d ha-1). Daily gains were similar (0.34 kg) for all germplasms. Bahiagrass stand declined rapidly for RRPS Cycle 14. Continuous stocking at high grazing pressure was not suitable for RRPS Cycle 14.
The proper assessment of defoliation by phytophagous insects in soybean [Glycine max (L.) Merr.] is required when screening germplasm for insect resistance. The objective of this study was to evaluate the effectiveness and relative efficiency of the widely used visual or relative assessment of insect defoliation compared with a more quantitative assessment for use in soybean breeding programs. In 1985 and 1991, visual assessments (whole plot, 1985; whole plant and average of partitioned plant, 1991) were compared with an established quantitative assessment. Twenty leaflets were randomly removed from each of the upper, middle, and lower thirds of the canopy of each plot. These leaflets were compared with a template of known defoliation in 1985. In 1991, a leaf area meter was used to determine the actual amount of defoliation of each plant. In 1990 and 1991, two types of visual estimates (whole plant and average of partitioned plant) were also compared with each other. Sampling time was estimated for each method. Estimates between visual and quantitative assessments were highly correlated each year. Precision was similar for each method. Sampling time for the visual assessment was at least 30 times faster than the quantitative assessment. The results of these studies show that relative estimates can be used to assess insect defoliation accurately and quickly.
Methods of evaluating forest damage are designed to provide information for detection, control, and preventative management. In this study the use of small-format aerial photographs to assess spruce budworm, Choristoneurafumiferana (Clemens), damage was evaluated. A procedure was developed to identify spruce–fir stand vulnerability and establish inventory data for silvicultural management considerations. Large-scale (1:4800) 35-mm color positives of eight spruce–fir stands were obtained in July and August 1979. Defoliation intensity and tree mortality of host species were the parameters used to assess damage severity. Individual stands were assigned separate damage rating values based on aerial and ground survey evaluations of defoliation and mortality. Comparison of aerial and ground damage severity ratings was accomplished with a contingency table. Results show photointerpretation of forest damage condition matched ground inspection in seven of eight stands surveyed.
Appel, D. N., Maggio, R. C., Nelson, E. L., and Jeger, M. J. 1989. Measurement of expanding oak wilt centers in live oak. Phytopathology 79:1318-1322. Local, tree-to-tree spread of Ceratocystis fagacearum was monitored live oak density. The largest initial focus had a lesser oak density and in live oak (Quercus fusiformis) with sequential, color infrared aerial increased from 1.5 to 6.6 ha, affecting 11,396 m 2 of crown cover. Crown photography during 1982-1987. A computer-based system for analyzing survival in 1987 ranged from 4 to 26% for trees that originally showed expanding foci was developed to measure rates of crown defoliation and symptoms in 1982. A strong linear correlation between the area of affected mortality, as well as expansion distances. Four foci expanded radially crown cover and total area occupied by each focus was noted. The rapid an average of 11-16 m/yr, with longer maximum distances of expansion rates of focus expansion were attributed to a high potential for root (up to 40 m/yr) commonly occurring. One focus that expanded most grafting and the occurrence of common root systems among clonally rapidly increased from 0.3 to 3.6 ha over 5 yr, affecting 10,774 m 2 of propagated live oaks. crown cover. This was initially the smallest focus, and it had the greatest Additional keywords: Ceratocystisfagacearum, epidemiology, Quercusfusiformis. Oak wilt "centers" are classical examples of disease foci, or in live oak; and to facilitate the placement of barriers for disease local concentrations of diseased plants about primary sites of control. This report describes the system and illustrates some infection (10). As defined, the size and shape of foci tend to important aspects of oak wilt epidemiology in Texas. influence patterns of further spread. Contaminated insects (Coleoptera: Nitidulidae) initiate foci in new areas by acquiring
Chrysophtharta bimaculata is the most economically important pest of eucalypt forests in Tasmania causing severe growth losses in plantations and seeded stands of the major commercial eucalypt species. An integrated pest management (IPM) strategy has been developed which allows peak activity of the natural enemies of C. bimaculata and monitors pest egg and larval populations before significant defoliation takes place. The relationship between beetle population level and degree of defoliation is quantified by regression models which are used to identify potentially damaging populations and the need for additional artificial control measures. The biotic insecticide, Bacillus thuringiensis var. tenebrionis was successfully tested against C. bimaculata larvae, and the use of B. thuringiensis and the synthetic pyrethroid, cypermethrin, in the IPM strategy is discussed.
Quantitative assessment of forest pest impacts is an important element in design of forest and pest management programs. Such assessment requires forecasts of pest populations, definition of the nature and extent of damage inflicted on trees by those populations, and translation of the damage effects across scales, from the tree to the stand to the forest. Central to this process are stand development forecasts which embody tree-level impacts of damage and which provide input to forest-level models. We discuss the role of stand growth forecasting in this context and propose a method for calibrating stand growth models that can be used to incorporate the effects of pest damage on tree and stand development. This calibration methodology is demonstrated for spruce budworm (Choristoneura fumiferana Clem.) effects on spruce (Picea sp.) and balsam fir (Abies balsamea (L.) Mill.) stands using permanent sample plot data acquired in New Brunswick over the past 20 years. The results showed liner relationships between tree diameter growth loss and cumulative defoliation, and non-linear relationships between reduced survival and cumulative defoliation. Growth loss relationships were similar for the species considered, while reduced survival relationships varied between species and age classes. Using these relationships as input to the STAMAN stand growth model, forecasts were made and compared against empirical studies of stand development under defoliation during spruce budworm outbreaks. The results suggest that reasonable approximations of stand response to pest incidence can be generated with relatively simple models and data sets. Judicious use of stand growth forecasts generated by such methods can help serve the needs of forest and pest management strategy design.
A general approach to regression modelling for ordered categorical response variables, y, is given, which is equally applicable to ordered and unordered y. The regressor variables x T =(x 1 ,···,x p ) may be continuous or categorical. The method is based on the logistic family which contains a hierarchy of regression models, ranging from ordered to unordered models. Ordered properties of the former, the stereotype model, are established. The choice between models is made empirically on the basis of model fit. This is particularly important for assessed, ordered categorical response variables, where it is not obvious a priori whether or not the ordering is relevant to the regression relationship. Model simplification is investigated in terms of whether or not the response categories are distinguishable with respect to x. The models are fitted iteratively using the method of maximum likelihood. Examples are given.
A stochastic simulation model was developed to derive a damage function for the spruce budworm, Choristoneura fumiferana Clem. (Lepidoptera: Tortricidae), and balsam fir, Abies balsamea (L.) Mill., herbivore interaction at the stand level for open-grown trees. Both aggregated and uniform models of attack pattern by late-instar larvae based on k of the negative binomial were evaluated to determine the impact of larval density and attack pattern on the loss in stemwood volume increment of young, thinned balsam fir stands in Newfoundland. Percentage loss in stemwood increment was a nonlinear, negative exponential function of initial larval density. Implementation of control measures to prevent 50% defoliation that is caused by about 14 larvae/branch tip would result in saving 24% of the annual stemwood increment after 1 year of defoliation and about 32% after 2 years of cumulative defoliation. Aggregation of spruce budworm larvae among trees within a forest stand results in less growth loss compared with a uniform pattern of attack. The nonlinear damage function may suggest tolerance and possibly compensatory growth after herbivory by low population levels of the spruce budworm.
Methods of evaluating forest damage are designed to provide information for detection, control, and preventative management. In this study the use of small-format aerial photographs to assess spruce budworm, Choristoneurafumiferana (Clemens), damage was evaluated. A procedure was developed to identify spruce–fir stand vulnerability and establish inventory data for silvicultural management considerations. Large-scale (1:4800) 35-mm color positives of eight spruce–fir stands were obtained in July and August 1979. Defoliation intensity and tree mortality of host species were the parameters used to assess damage severity. Individual stands were assigned separate damage rating values based on aerial and ground survey evaluations of defoliation and mortality. Comparison of aerial and ground damage severity ratings was accomplished with a contingency table. Results show photointerpretation of forest damage condition matched ground inspection in seven of eight stands surveyed.
Defoliation of eastern hemlock (Tsuga canadensis Carriere) forest caused mainly by the hemlock woolly adelgid (Adelges tsugae Annand) was detected, quantified, and mapped for a 1,267 km² study area in the New Jersey Highlands using anniversary dates of Landsat Thematic Mapper data (1984 and 1994). A model relating estimates of canopy condition to the temporal difference in near infrared/red reflectance (i.e., the vegetative index difference) was developed to predict and map four classes of hemlock condition across the study area. Data from 105 circular ground plots (90 m diameter) were used to develop the regression model, while data from 50 plots were reserved for accuracy assessment. The vegetative index difference was highly correlated to hemlock damage as measured on the ground (R² = 0.73). Lightly defoliated hemlock canopy did not differ spectrally from healthy hemlock; thus, these two classes were joined together. Accuracy assessment showed that hemlock condition can be predicted within one-half damage class with an overall accuracy of 64% for four damage classes, 70-72% for three classes, and 78-92% for two classes. Of the 7,735 ha of hemlock forest in 1984, 47% remained healthy to lightly defoliated, 44% had experienced moderate to severe defoliation, and 9% were dead by 1994. For. Sci. 43(3):327-335.
In Europe, crown transparency is one of the most important variables assessed for monitoring forest conditions. Though there are very strict international guidelines and intensive training programs for the evaluation of needle-leaf loss, there still exists a subjective component in the evaluation. This subjective component in the observation process can lead to a nonsampling error that can greatly reduce the precision of estimates of average needle-leaf loss. In this paper, methods for quantifying the potential loss in precision are presented and applied to data from the Swiss National Forest Health Monitoring Program. For. Sci. 41(4):758-776.
Functions are reported which predict percent budbreak and average leaf dry weight from elapsed degree-days (threshold = 4.4°C) for 6 important hosts of the gypsy moth (Lymantria dispar): Quercus alba, Q. rubra, Q. velutina, Fagus grandifolia, Acer rubrum, and A. saccharum. Budbreak observations are also summarized for Betula lenta and B. alleghaniensis. Day 105 (Julian date) was the best single date to start counting degree-days to predict percent budbreak for all species, years, and locations covered by this study. Simultaneous solution of the red oak leaf growth function and published gypsy moth larval growth and consumption functions predicted that an average larva will consume about 1,115 mg dry leaf weight, and drop about 156 mg. Apparent herbivory attributable to a single larva was calculated as 1,381 mg dry weight, where an average mature red oak leaf weighed 448 mg. Forest Sci. 29:607-617.
During feeding trials conducted on mature trees, consumption rate of red oak (Quercus rubra L.) foliage by gypsy moth (Lymantria dispar (L.)) larvae averaged 6.68 percent of average larval dry weight/degree day (threshold = 4.4°C). Average larval dry weight increased from 0.2 mg at hatch to 113.1 mg after 646.7 degree days spanning 43 days. The average accumulative consumption of red oak foliage was 1116.1 mg/larva during this period. Simple biological growth functions, which are useful in modeling defoliation, were developed to predict average larval dry weight growth and consumption. Forest Sci. 26:599-605.
Data from the annual assessments of forest condition in Switzerland indicate that there are a number of problems with the collection of even the most basic information relating to forest condition. Of the 13 indices examined, 3 are assessed as continuous variables and 10 as categorical variables. In all cases, significant (P < 0.05) differences were found in 1 or more years between the field teams and the control team(s). For two measures of defoliation, these differences were present in all years. The assessment of the categorical variables was more reliable, but the differences indicate that the data must be treated with caution, even when only two or three relatively simple options are available to the surveyors. The surveyors involved in the work were all experienced foresters who had received a high level of training. The results suggest that many data related to forest measurement and assessment that have been reported in the literature cannot be reliably compared. The results also indicate an urgent need for the intensification of quality assurance work (training and subsequent checking) within the context of forest assessment. For. Sci. 41(2):235-254.
The suitability of Garry oak (Quercus garryana) and red alder (Alnus rubra) as hosts for the gypsy moth, Lymantria dispar (L.), was assessed under laboratory conditions by observing larval survival, larval weights, foliage consumption, developmental period, pupal weight, and ova production. Survival was not significantly different between larvae fed Garry oak (98.7%) or red alder (97.4%). The mean maximum live larval weights were significantly different between sexes but not between diets. Females weighed 2,498 mg when fed Garry oak and 2,210 mg when fed red alder. Males weighed 894 mg when fed Garry oak and 737 mg when fed red alder. The mean amount of foliage consumed was significantly different between sexes but not diet. Female larvae consumed an average of 705 cm2 of Garry oak and 678 cm2 of red alder foliage. Male larvae consumed 247 cm2 of Garry oak and 253 cm2 of red alder foliage. The mean time from egg hatch to pupation was significantly different between sexes and diets. Female larvae pupated in 39.5 d on Garry oak and 48.1 d on red alder. Male larvae pupated in 33.4 d on Garry oak and 40.6 d on red alder. Pupal weights were not significantly different between sexes or diets. Male pupae weighed 554 mg from a larval diet of Garry oak and 572 mg from red alder. Female pupae weighed 1,846 mg from Garry oak and 1,711 mg from red alder. An average of 863 ova (Garry oak) versus 758 ova (red alder) was present in the reproductive tract of 2-d-old females, an insignificant difference. Pupal weights, frass production, and ova production were highly correlated. Nutritional indices indicated that Garry oak foliage was converted into biomass slightly more efficiently than that of red alder. These data indicated that foliage of either Garry oak or red alder provided a very suitable diet for the gypsy moth.
Ordinal regression techniques are introduced as a means to achieve multiple data analytical goals such as data smoothing for subsequent model derivation and prediction of response distributions conditional on some measurable covariates. Their usefulness to address some pertinent problems in forest biometry are discussed. For. Sci. 41(2):321-336.
A knowledge of the often complex relationship between insect populations and their effects on the yield-forming processes of crops is useful for assessing pest status and for devising methods of minimising the effects of infestation on yield. After a preliminary section on definition of terms, consideration is given to ways in which insects injure plants and crops, particularly annual crops. Investigations into the yield-forming processes of unattacked crops can often provide useful insights into the likely effect of insect injuries on yield. The differences between the characteristics of individuals and populations are discussed, both of insects attacking plants, and plants responding to the attacks of insects. The distribution of insects on plants and on crops and the effect that this has on damage is considered, as is the time and duration of the attack in relation to the growth cycle of the crop. The components of plant resistance to insect attack are then discussed, especially tolerance and the various ways in which plants and crops compensate for injury. The review concludes with an examination of the quantitative relationships between numbers of insects or injuries and their effects on yield. The generalised response is a sigmoid curve relating an upper and lower plateau of yield where changes in the numbers of insects or injuries have little effect. Most relationships conform to some part of this curve, though few show all its features.
Concentrating on material published from 1984-1987, examines 1) the impact of insect feeding on different aspects of plant performance, paying particular attention to flowering, fruit production, post-dispersal seed mortality, seedling mortality, defoliation, growth and reproduction, competitive ability, and death rate of established plants; and 2) reviews the impact of insect feeding on plant population dynamics, focusing on the release of specialist insect herbivores against target weed species in classical biological control projects, and the exclusion of insect herbivores from natural plant communities by repeated application of insecticides. -P.J.Jarvis
Little is known of the extent to which differences in methodology contribute to differences in reported levels of tree defoliation. In this study several common methods of assessing defoliation were compared on Eucalyptus blakelyi Maiden trees growing in a stand known to sustain a range of defoliation levels. Assessments were made of whole trees and of individual leaves collected at discrete intervals as well as of tagged samples of leaves that were monitored over a year. Despite fortnightly monitoring intervals, I was not able to identify with certainty any specific cause of loss of whole leaves other than senescence. Annual defoliation assessed on discrete samples of leaves was higher than annual defoliation assessed on monitored samples, presumably because the age of leaves in the discrete samples was sometimes underestimated. Leaf-level defoliation assessed from visual estimates was consistently higher than digitizer measurements of leaf areas, but calibration of estimates resulted in similar values. Values derived for tree-level defoliation from leaf-level measurements differed widely. depending on the method of calculation. For the samples collected from trees in this study, mean percentage defoliation per leaf was not an adequate descriptor of defoliation for the samples, since the frequency distributions of defoliation per leaf were markedly non-normal. This problem was avoided by using the samples, rather than the individual leaves, as the basic unit of expression of defoliation, with the total percentage defoliation per sample calculated from the ratio of total area of damage to total potential area, summed for all the leaves in the sample. For expanding leaves, the percentage of area damaged was not equivalent to the percentage of area missing from the expanded leaves, because of differences in the relative rates of growth of holes and leaves.
Abstract Field studies were conducted in 1995 to compare the photosynthesis response between simulated striped cucumber beetle, Acalymma vittatum (F.), defoliation and actual A. vittatum defoliation in ‘Carolina’ cucumber. Six simulated defoliation levels (0%–100%)were used over 5 timings of defoliation (first true leaf to harvest). Plots were defoliated with scissors twice each week throughout each timing interval, and defined as continuous defoliation. In addition to continuous defoliation, one-time simulated defoliation treatments were imposed using the same timing treatments but only 4 defoliation levels (0, 25, 50, and 100%). Two cage studies, with A. vittatum adults, were used to provide estimates of actual insect defoliation injury. Beetles were placed in cages for 2 weeks at densities of 0, 1, 3, 5, and 10/plant and net photosynthetic rates were measured. No significant differences (P <0. 05 were detected in pho-tosynthetic rates between simulated and actual insect defoliation treatments. Also, few significant differences in photosynthetic rates were detected between damaged and undamaged leaves in the simulated insect defoliation study. Results indicated that the simulated insect defoliation procedure appeared to accurately reflect the growth response of ‘Carolina’ cucumber subjected to actual A. vittatum defoliation, and that simulated defoliation results should be appropriate for developing economic injury levels for A. vittatum on cucumber.
Theory and recent literature suggest strong effects of induced plant defences in some plant herbivore systems. Few have studied behavioural effects on intact plants. Differences in foraging behaviour as well as weight gain were determined for first instar Helicoverpa armigera (Hübner) (Lepidoptera: Noctuidae) on Arabidopsis thaliana (L.) Heynh. (Brassicaceae) mutant and wild type plants, non-primed, or primed by herbivore feeding or methyl jasmonate. The differences in feeding were primarily in the length of feeding time as opposed to the area fed on, feeding location, or frequency. More larvae dispersed from plants after priming by mite feeding than dispersed after caterpillar feeding. Other behavioural activities such as resting were not significantly affected. Early instars gained less weight feeding on ein2 (ethylene insensitive) mutant, but there was no difference in weight gain between larvae feeding on induced and non-induced plants of the same type. We concluded that there are fitness consequences for neonates of the generalist H. armigera after feeding on induced plant tissues in some cases, and that distinct changes in behaviour are recognisable both at the fine scale and at grosser levels (dispersal). However, these changes are more subtle than might be expected.
Larval performance of the leaf beetle Paropsis atomaria Oliver was determined for larvae raised on both new and mature leaves of Eucalyptus blakelyi Maiden. Larvae were transferred to mature leaves at different ages; control larvae stayed on new leaves through all instars.
Only larvae reared on new leaves through the third instar survived to pupate on mature leaves; developmental time was prolonged by 20% and pupal weight was reduced by 50% in these larvae compared with larvae reared entirely on new leaves. Almost all larvae died when transferred to mature leaves as first, second or third instars.
Low survival and slow development on mature leaves was mainly due to failure by larvae to feed. Larvae palpated leaves and could discriminate among leaf ages immediately, without biting into the leaf tissue.
New leaves had higher concentrations of oil and tannins than old leaves, while there were no significant differences in nitrogen concentrations in the two types of leaves. Mature leaves were more than 3 times tougher than new leaves, in terms of g mm ⁻² of penetrometer force.
In drought years E. blakelyi may not produce sufficient new leaves to supply specialist herbivores with their preferred food resource. We infer that drought years reduce P. atomaria larval performance significantly, and influence the population dynamics of the insect.
Herbivores usually consume a mere fraction of available plant biomass. Spatial patterns in feeding damage may be attributable to induced defences by the host plant; a damaged plant reacts by lowering its nutritional value, thereby forcing herbivores to move on before food gets worse. In this study, we test this general hypothesis on a specific model system: caterpillars of the alpine butterfly Parnassius smintheus feeding on lance‐leaved stonecrop Sedum lanceolatum. We first describe spatial patterns in host distribution and feeding damage within alpine meadows. We then use laboratory experiments to test a key assumption behind the proposed mechanisms: that the host plant exhibits an induced response with a negative impact on larval performance, and that this response is activated with a delay. Finally, we relate the patterns observed to the actual behaviour of Parnassius larvae.
Overall, we found the level of feeding damage to be low (on damaged plants, only 5% of all leaves were fed upon). Within meadows, both host plants and feeding damage were clumped at a small spatial scale. This pattern seemed directly explicable by the timing of the host's induced defence. Laboratory experiments revealed a delay of 1–2 d before the defence reached a level affecting larval performance, and wild larvae switch plants more quickly than this. A simulation model demonstrated that the spatial distribution of host plant damage can be generated by a simple random walk, based on the empirically observed step frequency, length and turning angles. Hence, as the most parsimonious explanation for the observed level and pattern of host plant damage, we offer a scenario where induced changes in host‐plant quality limits the time spent per plant, but the herbivore moves throughout the landscape without any particular directionality.
Iwao's quadratic regression or Taylor's Power Law (TPL) are commonly used to model the variance as a function of the mean for sample counts of insect populations which exhibit spatial aggregation. The modeled variance and distribution of the mean are typically used in pest management programs to decide if the population is above the action threshold in any management unit (MU) (e.g., orchard, forest compartment). For nested or multi-level sampling the usual two-stage modeling procedure first obtains the sample variance for each MU and sampling level using ANOVA and then fits a regression of variance on the mean for each level using either Iwao or TPL variance models. Here this approach is compared to the single-stage procedure of fitting a generalized linear mixed model (GLMM) directly to the count data with both approaches demonstrated using 2-level sampling. GLMMs and additive GLMMs (AGLMMs) with conditional Poisson variance function as well as the extension to the negative binomial are described. Generalization to more than two sampling levels is outlined. Formulae for calculating optimal relative sample sizes (ORSS) and the operating characteristic curve for the control decision are given for each model. The ORSS are independent of the mean in the case of the AGLMMs. The application described is estimation of the variance of the mean number of leaves per shoot occupied by immature stages of a defoliator of eucalypts, the Tasmanian Eucalyptus leaf beetle, based on a sample of trees within plots from each forest compartment. Historical population monitoring data were fitted using the above approaches.
Because the tip of many dicot leaves matures and ceases expansion well before the base, we predicted that the removal of a given amount of leaf tissue from the base of an expanding leaf would result in greater reductions in final leaf area and overall plant performance than removal of the same amount of tissue from the tip of an expanding leaf or from either the base or tip of mature, fully expanded leaves. We tested this notion by removing a circular 3.9 cm2 hole from either the base or tip of rapidly expanding leaves (20–30% expanded, two nodes from the apex) or nearly fully expanded, mature leaves (85–100% expanded, five nodes from the apex) of tobacco (Nicotiana tabaccum) and measuring the final area of the hole, the final area of the fully expanded damaged leaf, and the number and mass of fruits produced by a plant. A given amount of area removed from the base of an expanding leaf resulted in almost 4 times the amount of visible damage than occurred when the same amount of damage was applied to the tip or base of a mature leaf and over twice the amount of visible damage than occurred on the tip of an expanding leaf. Furthermore, damage to the base of an expanding leaf resulted in nearly a 40% reduction in the final area that the leaf would have achieved without damage and a 35% reduction in the number and mass of fruits produced. These results not only suggest that where on a leaf a folivore feeds has consequences to the ultimate area that a leaf can reach and to overall plant performance, but they also have strong implications for a number of research areas in plant-herbivore interactions. For example, these data show that a lack of consideration of leaf developmental patterns can result in gross overestimates of consumption by folivores and severe under-estimates of the effect of folivory on leaf area display.