Drought Effects on Damage by Forest Insects and Pathogens: a Meta-Analysis

Climate change may decrease summer precipitation and increase winter precipitation across the Northern Hemisphere, resulting in dire consequences for forest ecosystems since summer drought damages tree growth and forest ecosystem functioning. In addition, prolonged droughts may also trigger more frequent or severe outbreaks of forest insects and pathogen epidemics, and these events could interact with carbon starvation or hydraulic failure to further increase rates of tree mortality. Moreover, there is considerable variation in the magnitude and direction of responses to water stress by pathogens and insects. Therefore, to draw general conclusions about tree drought-damage, insect-damage, and pathogen-damage relationships, Jactel et al. (2012) conducted a meta-analysis of published primary studies that addressed the impact of water stress on forest pest and pathogen damage. More specifically, the authors estimated the overall effect of water stress on insect pest and fungal damage in forest trees and investigated the variation of response to water stress among functional groups of pests and pathogens. They also explored the relationship between the magnitude of pest or pathogen damage and the severity of drought. Jactel et al. found that primary damaging agents living in wood caused lower damage to water-stressed trees, while primary pests and pathogens living on foliage caused more damage to water-stressed trees, in all cases irrespective of stress severity. Damage by secondary agents increased with stress severity. Overall, insect and fungus feeding behavior, affected tree part, and water stress severity were the three main predictors of forest damage in drought conditions.—Megan Smith
Jactel, H., Petit, J., Desprez-Loustau, M.L., Delzon, S., Piou, D., Battisti, A., Koricheva, J. 2012. Drought Effects on Damage by Forests Insects and Pathogens: a Meta-Analysis. Global Change Biology 18: 267 – 276. DOI: 10.1111/j.1365-2486.2011.02512.x

Numerous textbooks have described the variable responses of forest pests to tree water stress, most of which is related to insect feeding guild. Generally, bark beetles and woodborers perform better under severe drought scenarios, while sapsuckers also benefit from water-stressed trees under moderate drought conditions. The effect of drought on leaf miners, leaf chewers, and gall makers is uncertain. Pathogenicity may also be enhanced or reduced with increased drought. Furthermore, the duration and severity of water stress influences insects’ and pathogens’ responses to drought. One scientific study found that infections are more likely to develop during or after prolonged drought stress.
To draw general conclusions about the diverse drought-damage relationships between pathogens, insects, and trees, the authors conducted a meta-analysis of published primary studies that investigated the impact of water stress on forest pest or pathogen damage. Meta-analysis is a set of statistical tools that combines the outcomes of independent studies to evaluate the overall effect of a particular factor. It also tests the influence of covariates on this effect.
For their meta-analysis, the authors collected published studies that compared pest or disease damage on water-stressed vs. control trees. Studies were included in the analysis if they met specific criteria. One criterion stated that the study must have assessed tree damage caused by an insect or fungal pathogen. Damage variables included measures that quantified impact on tree survival or tree growth by recording the amount of damaged or consumed tree tissues, the number of attacks per tree, or the percentage of infested or killed trees. Studies were also included if they reported any insect and fungal species that affected tree tissues or organs. The second criterion stated that the mean response variable (tree damage), a measure of the variance, and the sample size for both control and drought treatments must have been reported to be included in the meta-analysis. The third criterion affirmed that water conditions in the control and stressed group of trees must have been quantified using predawn leaf water potential with a pressure chamber. This ensured that two groups of trees were under different water supply conditions and that the methodology of water stress assessment was consistent across studies. Predawn leaf water potential values were used as indicators of water stress severity. Finally, the fourth criterion stated that the reported paired comparison between water-stressed and unstressed (control) trees must have been made under the same environmental conditions (besides water supply), on the same date and in the same area.
The effect of water stress on forest insect and disease damage was estimated by calculating Hedges’ das a measure of the effect size. A positive dvalue indicated higher damage on water-stressed trees than on control trees. The authors selected the one variable per comparison between water-stressed and unstressed trees that had the largest sample size or allowed the highest number of paired comparisons. Data was only used from the first year of the studies and only data from the first application of water treatments on the trees were utilized. If results were reported for 2 years but from two different, independent tree samples, data for each year were used as two separate comparisons.
Jactel et al. quantified water stress severity with four variables. The first two were calculated with the information provided in the retained papers as the difference or ratio between the mean predawn leaf water potential in water-stressed and control trees. The higher the absolute value of predawn leaf water potential, the more water-stressed the tree was. The other two variables represented the hydraulic failure in the trees. These variables were calculated as the difference or ratio between the mean predawn leaf water potential in water-stressed trees and the xylem (the vascular tissue in plants that conducts water and nutrients upward from the root) pressure inducing 50% loss in hydraulic conductance (P50) due to cavitation in the same tree species. P50, as a representation of cavitation resistance, is highly variable between species and correlated with plant drought tolerance (lethal water stress).
The authors split the dataset into subsets of different functional groups of insects or fungi depending on their feeding substrate, and separated insects or fungal species colonizing foliar organs involved in photosynthetic processes (leaves, needles) vs. those living in woody organs responsible for tree structure (bark, wood, roots). Then, they distinguished between insect and fungal species that develop on healthy trees (primary agents) and those that utilize trees in poor physiological conditions (secondary agents). Jactel et al. assumed that there would be four functional groups, but they were unable to find examples of secondary agents that damage foliar organs. Therefore, the study only included three functional groups. Species were also classified based on their trophic guild: chewing, boring, sucking and galling insects, leaf pathogens, root and bark rot, blue-stain fungi, and endophytes.
Studies were categorized as observational or experimental depending on whether the drought was caused by natural conditions or controlled water supply. Additionally, the authors distinguished between comparisons made in the field and those made in protected conditions in absence of natural enemies.
Effect sizes across all comparisons were combined using the random effects model to yield the grand mean effect size (d++). The effect was considered significant if the bootstrap confidence interval, calculated with 9999 iterations, did not include zero. The mean effect size (d+) and 95% bias-corrected bootstrap confidence interval were calculated for each functional group of forest insect or fungi combining affected tree organ and its physiological status. A mixed effect model was used to test the between class heterogeneity and to test for significance of class effect. A P value of 0.001 was used to test for statistical significance. A mixed model was also used to test the relationship between the differences in damage on stressed vs. control trees (effect size) and severity of water stress (continuous variable).
Finally, the authors calculated a fail-safe sample size that represented an estimate of the number of non-significant, unpublished, or missing studies that would need to be added to the analysis to make the overall test of an effect statistically non-significant. After, testing, the authors found that their results were unlikely to be affected by publication bias.
Jactel et al. derived 100 comparisons of forest pest and disease damage on water-stressed vs. unstressed trees from 40 publications and reports. They involved 27 insect and 14 fungus species. A total of 26 tree or shrub species were studied. Overall, water stress resulted in higher forest pest and disease damage. The grand mean effect size equaled 0.23 and was significantly different from zero. However, an effect size of 0.2 is considered a small effect. Additionally, 40% of the individual effects were also negative, indicating lower damage in water-stressed trees than un-stressed trees. A graph displaying Hedges’ d effect size of 100 individual studies was constructed. Negative effect sizes indicated that drought resulted in lower damage.
The type of trophic substrate used by forest pest and pathogens had a highly significant effect on the difference in damage between water-stressed and unstressed trees. Primary damaging agents living on foliar organs caused higher damage in water-stressed trees than un-stressed trees, irrespective of stress severity. Drought did not exacerbate damage caused by primary agents that developed on woody organs. However, it did increase damage caused by secondary agents that developed on woody organs. A table displaying the mean Hedges’ effect size per functional group of forest pest and pathogen was constructed.
The effects of trophic guild were never significant within each functional group of forest pests and pathogens. Damage caused by sucking and boring insects and root and bark rot fungus species developing in woody organs in healthy trees (primary agents) was not worsened by drought. Drought resulted in slightly higher damage caused by leaf pathogens living on foliar organs in healthy trees and galling and chewing insects. Additionally, endophytic fungi damage increased with drought, but the results were not statistically significantly different from zero for boring insects and blue-stain fungi. Contrary to previous studies, these results suggest that the effect of water stress on the level of damage by forest pests and pathogens depends more on the type of substrate they use rather than on their feeding guild. A table displaying the effects of drought on mean effect size (damage) by different types of forest pest and pathogens was constructed.
After testing the effect of the type of water stress application on level of damage for each functional group of forest pests and pathogens separately, the authors found that there was no significant difference in mean effect size between observational and experimental studies. There also was no difference between studies made in the field or in protected conditions.
Finally, Jactel et al. tested the effect of water stress severity on level of damage for each functional group of forest pests and pathogens separately. Water severity did not affect the level of damage in water-stressed trees for any primary damaging agent. However, water stress severity did affect the level of damage caused by secondary agents living in woody organs. The variable that best explained damage variation was the ratio between observed predawn leaf water potential in the stressed trees and the species-specific index of drought tolerance (P50). Damage was consistently higher in stressed trees with a predawn leaf water potential higher than 30% of P50. Below 30%, water-stressed trees were more likely to have less damage than unstressed trees. Both secondary fungi and insect species living in woody organs similarly affected the water-stressed and unstressed trees (there was no significant difference between the two groups). A graph displaying the relationship between levels of damage (effect size) made by secondary forest pests and pathogens living on woody organs and water stress severity was constructed.
Previous studies have found that insect performance response to water stress depends on their feeding traits. Borers are known to perform better on stressed plants while gall makers and leaf chewers are negatively affected. However, one study found that sapsuckers benefited from drought while another study’s results contradicted these findings. Furthermore, drought is thought to negatively affect forest pathogens because fungi require high humidity conditions for spore dispersal, germination, and infection. Overall, the combination of effects on the performance of the biotic agent and effects on tree response could explain discrepancies in the results between different studies.
Drought can affect the nutritional quality of host trees for herbivorous insect and fungal pathogens through changes in water, carbohydrates, and nitrogen contents. Water supply greatly influences carbohydrate photosynthesis and therefore the provision of sugars for insects and parasitic fungi. As a result of drought, the reduced concentration of carbohydrates in conifer bark tissues may reduce the development of bark beetles and blue stain fungi. Furthermore, reduced water content and protein hydrolysis lead to higher nitrogen concentration in tree organs during drought. Since nitrogen is a limiting nutrient for many insects, an increase in plant nitrogen during water stress could improve the performance of phytophagous insects. For example, defoliator performances are higher in moderately water-stressed trees due to higher concentration of soluble nitrogen in foliage. Sap feeding insects also benefit from an increase in nitrogen.
Moreover, amino acids can be found in increased concentrations in water-stressed trees, stimulating the growth of bark canker fungus. As a result, the concentrations of carbohydrates and nitrogen decrease in the stem of the tree under moderate stress. This would limit the performances and then the damage of primary pests living on woody organs. Performance and damage made by primary pests living on foliar organs (which benefit from higher nitrogen content) would increase.
Water-stress also affects host metabolism involved in resistance to pest and pathogen damage. Tannins utilized in tree resistance can be found in higher concentrations in foliage of water-stressed trees, which deter leaf chewers such as beetles and lepidopteron. In contrast, resistance mechanisms may also be less effective in water-stressed trees.
Lower water supply affects sap flow and oleoresin production and pressure, which results in lower resistance to primary attacks of many bark beetles. Infection of pathogenic blue-stain fungi by scolytids leads to the development of necrotic lesions containing concentrations of terpenoid and phenolic chemicals that are toxic to insects and fungi. However, water-stressed trees lack the carbohydrates reserve to fuel the secondary metabolism involved in these resistance processes. Therefore, severely water-stressed trees are likely to be more damaged by secondary pest and pathogens like wood boring insects and blue-stain fungi.
Conversely, moderate water stress could lead to increased resistance. Because a tree’s carbohydrate pool still increases under moderate water stress, the tree may instead allocate its carbohydrates to the synthesis of defensive secondary chemicals rather than to growth and development. Secondary pests living in woody organs, like bark beetles, would then cause less damage in moderately water-stressed trees.
Additionally, decreased water content in severely stressed trees could lead to tougher foliage, resulting in lower herbivory by chewing insects.
Overall, the authors’ results confirm that drought does not systematically result in higher biotic damage. Two factors that explained the tree damage response to drought were type of feeding substrate for forest insect and pathogens and water stress severity. 

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s