Summary

The key of survival for agronomic weeds is their adaptability to agronomic conditions, which are marked by regular disturbances. Modern agricultural production is defined by short crop rotations with a share of winter crops, the application of strongly selective herbicides and a weed flora that suffers from a loss of diversity. Weeds, especially grasses, have nevertheless been able to profit from these developments. In the last 30 years, Apera spcia-venti (L.) P. Beauv. has emerged as one of the most important grass weeds in northern, middle and eastern European crop production. The range and density of the winter annual has distinctly increased. Additionally, A. spcia-venti has evolved resistances to multiple herbicides with varying modes of action. Biological characteristics, such as wind pollination, high rate of seed production per plant and high genetic variability, appear to have predestined the species for this evolution.
The aim of this work consisted of an analysis of the population dynamic and development of herbicide resistance of A. spcia-venti.
The portion of winter cereals in the crop rotation and the nitrogen fertilizer application rate affect the growth of A. spcia-venti. A measurement of the influence is the tillering rate. This was significantly higher in the crop rotation composed of 50 % winter cereals than in the crop rotation composed of 67 % winter cereals. Reducing the nitrogen fertilizer application rate by half lead to a significant increase in the tillering rate.
The investigations into the population dynamics of A. spica-venti are based on data from a 15-year-long field experiment that ran at the University of Göttingen from 1981 to 1996. The development of A. spica-venti was recorded in the crop rotation beets – winter wheat – winter barley. It is an exceptional case, that data over a period of 15 years or five complete cycles of a crop rotation were available for the analysis and stochastic estimation of all population dynamic parameters. In the cereals A. spica-venti was able to grow long enough to produce seeds, whereas seedling were completely managed in the beets. Along with the plant density, the soil seed bank was sampled and the seed production of individual plants were determined on three occasions each year. This collection of data has been analyzed for the first time in the work presented here.
Analysis of population’s development showed that when starting with a small soil seed bank, it took three cycles of the crop rotation before a significant increase occurred. Compared to the first crop rotation, the median number of seeds in the soil seed bank had increased 9.5-fold by the end of the 15-year observation period. The development of the plant density did not coincide with the soil seed bank; there was no significant increase. No difference in the development of A. spica-venti between the two cereals was evident. The later harvest of winter wheat favoured the ripening of A. spica-venti and the introduction of new seeds into the population.
Based on the data analysis, population dynamic parameter values for A. spica-venti were estimated with linear mixed models. The method offers the benefit of incorporating the yearly variations as well as the structure of the experimental design. The entire life cycle of A. spica-venti was quantified individually for beets, winter wheat and winter barley. On the basis of this, a stochastic population dynamics model for A. spica-venti could be developed. Starting with a known supply of seeds in the soil, the model can simulate the development of A. spica-venti in the abovementioned crops. The ability to apply the modelt o other crops and crop rotations was tested. To this end, data from another location with differing crops and crop rotation was available. In a comparison of the simulated and real plant densities in spring, the model proved to underestimate this result. In contrast, agreement was achieved for the panicle density before harvest.
Weed populations conform to the special conditions of the locations where they are found. Regularly herbicide applications belong to this for many years. Field populations are tested in bioassays for herbicide resistance. Bioassay results may be influced by general conditions. In regard to this, it is important that the test plants are in a similar stage of development when sprayed. The germination behavior of various A. spica-venti populations were examined under differing temperatures. Populations that originated from conventionally-managed locations and were resistant to isoproturon germinated significantly quicker and more consolidated than sensitive populations from organically-managed locations.
Bioassays conducted on whole plants in greenhouses offers the possibility to make a first estimate of the degree of resistance independently of the resistance mechanism. With the aid of a reference population, the degree of resistance of the tested field population(s) can be estimated. Therefore, the origin of the reference population is important. As a first step, the susceptibility of A. spica-venti reference populations from various test laboratories to iodosulfuron were tested. The differences in susceptibility were minimal. Nonetheless, the origin and cleanliness of the seed played a role in the results: well-cleaned seed from certified seed distributors reacted with less variability than seed harvested from field populations.
Often suspected resistance in a population cannot be proven. Bioassays are indeed simple to execute, but very space and work intensive. It would be advantageous, if selected sample size could be reduced from the outset using the distribution of plants within a field (position and plant density) without overlooking early resistance developments. Also, experiences and assessments by the farmers could be indicators. Two studies, in one field and between different fields, were carried out in order to prove these options.
Within one field A. spica-venti seeds were sampled from plants grown in various density categories (single plants, sparse and dense) and subjected to dose response tests with isoproturon in ordert o identify differences in susceptibility. The ED50 values within the one field varied significantly. The efficacy of the herbicide was not dependent on density. But, there was a spatial association with the processing direction in the field.
The second investigation was a two-year-long monitoring to identify the resistance status of A. spica-venti. Seeds from 38 winter wheat fields in northern and eastern Germany were tested. Additional information about the fields was obtained from a questionnaire. This information included management practices of the field and if the farmer suspected resistance in the field.
The evolution of resistance was tested against three modes of action by means of five active ingredients (ACCase-inhibitor with clodinafop, ALS-inhibitor with flupyrsulfuron and iodosulfuron and photosystem-II-inhibitor with chlorotoluron and isoproturon).
In 15 of the 38 populations resistance was confirmed to at least one herbicide. Resistance was associated with a soil management regime that switches between ploughing and never ploughing as well as with normal sowing times. 14 farmers were correct in their assumptions regarding absence of herbicide resistance in their fields. However, resistance was only proven in half of the fields where farmers suspected resistance based on high densities of A. spica-venti.
Resistance can be inherited mono- or polygenetically. How this plays out in the population development in the field can be simulated with models. The model PERTH (Polygenic Evolution of Resistance to Herbicides) is an individually-based, stochoastic simulation model developed just for this purpose. This model was applied in the current work for A. spica-venti in the crop rotation beets – winter wheat – winter barley and modified with the corresponding population dynamic parameters. Simulation studies for the population development over 15 years were conducted under the presumption that four resistance-inducing alleles are present. Duplication of the frequency of the resistance-inducing alleles in the first 10 years lead to purely susceptible genotypes being removed from the population and an increase in the occurrence of purely resistant genotypes. Increasing plant densities in the field was the consequence since the efficacy of the herbicide was no longer sufficient.
In the present work a contribution to understanding the population dynamic and herbicide resistance development of A. spica-venti was made.
The investigations demonstrated above all that the development of A. spica-venti is variable and therefore difficult to predict. The plant density remains at a similar level over many years while the soil seed bank subtly grows to a significant amount. Unexpectedly high plant densities after herbicide applications are occasionally rather a result of environmental situations than resistance. The investigations also showed that management lies completely in the hands of farmers. Weed seeds can enter a field during handling of the headland, consequently inducing differences in susceptibility to herbicides. Important seems to be regular follow-up checks of fields after herbicide applications since A. spica-venti is particularly identifiable in fields with its erect form and prominent panicles.