Innate and introduced resistance traits in genetically modified aspen trees and their effect on leaf beetle feeding.

Genetic modifications of trees may provide many benefits, e.g. increase production, and mitigate climate change and herbivore impacts on forests. However, genetic modifications sometimes result in unintended effects on innate traits involved in plant-herbivore interactions. The importance of intentional changes in plant defence relative to unintentional changes and the natural variation among clones used in forestry has not been evaluated. By a combination of biochemical measurements and bioassays we investigated if insect feeding on GM aspens is more affected by intentional (induction Bt toxins) than of unintentional, non-target changes or clonal differences in innate plant defence. We used two hybrid wildtype clones (Populus tremula x P. tremuloides and Populus tremula x P. alba) of aspen that have been genetically modified for 1) insect resistance (two Bt lines) or 2) reduced lignin properties (two lines COMT and CAD), respectively. Our measurements of biochemical properties suggest that unintended changes by GM modifications (occurring due to events in the transformation process) in innate plant defence (phenolic compounds) were generally smaller but fundamentally different than differences seen among different wildtype clones (e.g. quantitative and qualitative, respectively). However, neither clonal differences between the two wildtype clones nor unintended changes in phytochemistry influenced consumption by the leaf beetle (Phratora vitellinae). By contrast, Bt induction had a strong direct intended effect as well as a post experiment effect on leaf beetle consumption. The latter suggested lasting reduction of beetle fitness following Bt exposure that is likely due to intestinal damage suffered by the initial Bt exposure. We conclude that Bt induction clearly have intended effects on a target species. Furthermore, the effect of unintended changes in innate plant defence traits, when they occur, are context dependent and have in comparison to Bt induction probably less pronounced effect on targeted herbivores.

Increased resistance of Bt aspens to Phratora vitellinae (Coleoptera) leads to increased plant growth under experimental conditions.

One main aim with genetic modification (GM) of trees is to produce plants that are resistant to various types of pests. The effectiveness of GM-introduced toxins against specific pest species on trees has been shown in the laboratory. However, few attempts have been made to determine if the production of these toxins and reduced herbivory will translate into increased tree productivity. We established an experiment with two lines of potted aspens (Populus tremula×Populus tremuloides) which express Bt (Bacillus thuringiensis) toxins and the isogenic wildtype (Wt) in the lab. The goal was to explore how experimentally controlled levels of a targeted leaf beetle Phratora vitellinae (Coleoptera; Chrysomelidae) influenced leaf damage severity, leaf beetle performance and the growth of aspen. Four patterns emerged. Firstly, we found clear evidence that Bt toxins reduce leaf damage. The damage on the Bt lines was significantly lower than for the Wt line in high and low herbivory treatment, respectively. Secondly, Bt toxins had a significant negative effect on leaf beetle survival. Thirdly, the significant decrease in height of the Wt line with increasing herbivory and the relative increase in height of one of the Bt lines compared with the Wt line in the presence of herbivores suggest that this also might translate into increased biomass production of Bt trees. This realized benefit was context-dependent and is likely to be manifested only if herbivore pressure is sufficiently high. However, these herbivore induced patterns did not translate into significant affect on biomass, instead one Bt line overall produced less biomass than the Wt. Fourthly, compiled results suggest that the growth reduction in one Bt line as indicated here is likely due to events in the transformation process and that a hypothesized cost of producing Bt toxins is of subordinate significance.

The effect and transmission of one isolate of the rust Puccinia minussensis on five clones of Lactuca sibirica.

In a greenhouse experiment, one isolate of the systemic rust fungus Puccinia minussensis was applied to the host clone from which it was collected and to four other clones of the host Lactuca sibirica. The plants were grown in fertilized potting compost (N+) to promote growth and in peat (N-) to hamper growth, for three growing periods during one year. The results show that the expression of host plant resistance could not be determined visually, but there were differences in effects on the clones. The rust isolate was found to produce a significantly higher percentage of diseased shoots on clone A (the clone it was taken from). Furthermore, the rust also had the strongest effect on both biomass and shoot production on clone A compared to the other four clones. The data suggest that the rust isolate is highly adapted to the clone from which it originated. We suggest that selection in this system has not favoured a benign pathogen and that similar patterns are likely to occur for plants that (i) rarely establish by seeds; (ii) have strong lateral growth; and (iii) may persist for long periods once established.

Density-dependent mortality in Pinus sylvestris caused by the snow blight pathogen Phacidium infestans.

Relationships between disease incidence and the density of host plant populations were investigated in the Pinus sylvestris-Phacidium infestans host-fungal pathogen association, in which the season of death of plants killed up to 3 years previously could be accurately determined. Significant (P<0.05), positive density-dependent relationships between the proportion of plants dying in the winters of 1987-1988, 1988-1989 or 1989-1990 and the original stand density were detected in 12 of 26 comparisons. Of the remaining comparisons, all but three had positive regression coefficients for the same association. Plants killed up to 2 years previously contributed to inoculum production. The use of "standing dead" as a predictor in the analyses showed that the proportion of plants dying in the winters of 1988-1989 or 1989-1990 was generally better correlated with standing dead in the previous summer than with the density of the original population. Significant (P<0.05), positive density-dependent associations were also found between the proportion of living plants in 1990 infected with P. infestans and the number of standing dead plants in all nine comparisons. In contrast, only four of the nine associations between these proportions of infected plants and population density were significant. The strength of the density-dependent relationships varied substantially within and between sites. Much of this variation appears to be due to differences in the stage of development of the epidemics occurring at different sites.

The interaction between the clonal herb Trientalis europaea and the host specific smut fungus Urocystis trientalis.

The interaction between the clonal dicotyledonous herb Trientalis europaea and the systemic smut fungus Urocystis trientalis was investigated. By marking individual plants in the field and transplanting plants to the greenhouse, disease transmission and the effect of disease on survival and fecundity of plants were estimated. Field data showed that 50% of the diseased and none of the healthy plants died during summer. Surviving diseased plants produced significantly fewer winter buds than healthy plants (means ±S.E. 1.12±0.05 and 1.88±0.07, respectively). Seed capsule production was low overall and did not differ between diseased and healthy plants. Disease was not seed-transmitted and transmission from infected mother plants to daughter ramets was not total (means 33% and 46%, in two experiments). Disease transmission was also influenced by light conditions.