Search

Reversing the decline of biodiversity in European agricultural landscapes is urgent. We suggest eight measures addressing politics, economics, and civil society to instigate transformative changes in agricultural landscapes. We emphasize the need for a well-informed society and political measures promoting sustainable farming by combining food production and biodiversity conservation.

Glyphosate herbicides play a prominent role internationally in weed management due to their broad spectrum effectiveness. There are concerns that a high reliance on glyphosate herbicides will negatively affect biodiversity over time. In Europe, where cropping systems are carried out without glyphosate resistant crops (GMO), glyphosate herbicides are applied between cropping seasons either in combination with or in place of plowing. We explored the effects of mouldboard plowing, chisel plowing, and glyphosate herbicide application (in a dose reflecting 100% of the manufacturer’s recommendation and a dose reflecting 50%) on weed species number, density, true diversity and community composition. The field experiment was conducted for three consecutive years in a randomized plot design on a naturally occuring weed community without a crop. A total of 83 species and 4 genus were identified over the three years of surveying. By applying a Multinomial Species Classification to the dataset, it was determined that root and rhizome propagating species as well as annual agricultural weeds with no pronounced seasonality in their germination were rarer on glyphosate treated plots than on tillaged plots. Promoted by glyphosate application were in particular grassland legumes and hemicryptophytes. Partial Redundancy Analyses (pRDA) showed that the species communities evolved over the years with dependency on the type of treatment and timing of the treatments. The overall average number of species was 19.7/m2 and the true diversity was 8.3/m2. The average density was 295 plants/m2. In all of the treatments, a factorial ANOVA demonstrated that species numbers and true diversity sank in all three of the years of the experiment over the course of the vegetation period whereas the density tended to rise. The reality of weed management is that any method employed will influence the weed composition in some way. Our experiment demonstrated that some species will be favored over others depending on the weed management method, but the overall biodiversity of the weed community was not more negatively affected by one method compared to another.

Renewable energy produced at biogas plants has tremendously increased the demand for digestible biomass in central Europe over the last twenty years. As a consequence, new crop rotations combining conventional and recently introduced bioenergy crops are tested in long-term field trials. The aim of this work was to determine risks for soilborne wheat stem base diseases in new bioenergy crop rotations. Crop rotations tested consisted of maize, sorghum, cereals, grasslegume mixtures, and phacelia. Field trials with five rotations were established at four different sites in Germany, whereby winter wheat was grown at the end of each rotation. Field trials were set up in a block or strip plot design in 2005 and 2006. After eight years, stem base diseases of winter wheat were visually assessed during milk ripeness. ‘Random forests’, a non-parametric statistical method, was used to select the most important variables describing weather, soil productivity and rotation. Subsequently, disease incidences were modeled with generalized linear models. Disease incidence was high for Fusarium foot rot (47%), moderately high for eyespot (30%), and low for sharp eyespot (4%). An increasing number of days with high precipitation during six weeks after sowing, and less growing degree days from February through June led to higher eyespot incidence (R2 = 60%). Fusarium foot rot incidence was correlated to high mean precipitation during six weeks after sowing, and low number of growing degree days from February through June (R2 = 78%). In contrast to the eyespot model, year and region had additional explanatory power to the Fusarium foot rot model. Based on the results of this study we could show that annual weather patterns had a high impact on eyespot and Fusarium foot rot incidences, while the effect of rotation was low. Although unexplained site and year-specific effects require further investigations, weather patterns identified by applying an innovative variable selection method provide a valuable starting point for developing and improving forecasting models for Fusarium foot rot.

We present a detailed analysis of glyphosate-based herbicide (GBH) use patterns with a dataset of herbicide use on arable farms in five districts of Northern Germany. Two aspects are focused: (1) The relationship between GBH and selective herbicides used on a specific field in the same cropping period and (2) the interplay of tillage, GBH and selective herbicide use in weed management. The most interesting result of our analysis is that the intensity of selective herbicide use for in-crop weed management was not related to GBH treatment prior to the crop (except maize). GBH treatments increased the total herbicide intensity on a field. Hence, using GBH pre-crop does not necessarily save herbicides in the following crop. GBH were additional treatments, not replacing selective herbicides. Our data was collected from different districts in Northern Germany extending from West to East. In the Eastern districts, a larger proportion of fields were treated with GBH. When a field was treated with GBH, it was applied with a similar intensity, regardless of the district. Hence, differences in GBH use patterns were caused only by the decision to treat pre-crop, while the intensity of these GBH treatments is hardly varied. In the Eastern districts, the higher proportion of fields treated with GBH is associated with larger proportion of fields with non-plough tillage. Non-plough fields were also treated more intensively with selective herbicides. Nevertheless, the use of both GBH and selective herbicides on the same field was more frequent and intensive in the Western districts of Northern Germany. We attribute the differences in pre-crop GBH use and in-crop selective herbicide use to regional production differences. Small to medium proportions of fields with both ploughing for primary tillage and pre-crop GBH use in all districts and most prominent in winter barley, winter wheat and maize indicate a luxurious, easy reducible use of GBH. We propose to promote farms with non- and low-use of GBH as an example of good practice.

Weeds typically occur in patches of different sizes and densities. Post-dispersal seed predators contribute to biological weed control by removing newly produced weed seeds before they can enter the seedbank. Seed predation can limit weed population growth if seed predators respond to seed densities so that seed predation is highest within weed patches. To determine whether this relationship holds for Echinochloa crus-galli seeds in corn fields, we observed the levels and responses of seed predators to different densities of E. crus-galli seeds during autumn in 2014 and 2015 (August until corn harvest in September–October) and during the winter of 2014 (August 2014 until March 2015) in three corn fields. We tested whether seed predation in autumn is a good estimate of seed predation from seed shed until the following spring (autumn plus winter), because seed predators were assumed to be less active during winter. In autumn 2014, the levels of seed predation and the response to seed density differed between fields because the main type of seed predator varied between fields. In autumn 2015, almost all seeds were lost due to seed predation in all fields, probably because of a high abundance of rodents. A response to seed density could not be detected. Seed predation during winter resulted partially in an increased level and a density independent response to seed density in all fields. Seed predation in autumn does not reflect seed predation from seed shed until the following spring. Furthermore, the effect of seed predation on weed population dynamics will differ between years and fields.

The reduction of pesticide use intensity is a societal and political ambition. Crop rotation is one important method to control pests and diseases in arable farming. We investigated the contribution of crop rotation to the variability of herbicide and fungicide use of 60 farms in four regions of Northern Germany. Our study aimed at answering the question: do diverse crop sequences lead to reduced herbicide and fungicide use in arable farming? Ten-year data on chemical plant protection measures and field management were examined for six field crops. We classified crop sequences (triplets of three succeeding crops) according to their susceptibility for weeds and diseases (= ’riskiness’). The Treatment Frequency Index (TFI) of the last crop in the triplet was set in relation to the crop triplet riskiness, additionally also in combination with tillage. In general, herbicide and fungicide use intensities were smaller in more diverse crop sequences. Diversified cereal sequences, involving roots and tubers, maize or spring cereals were less dependent on herbicides. Cultivation of maize in three subsequent years increased herbicide use. Crop sequences including high proportion of winter cereals increased fungicide use in cereals, while roots and tubers, winter oilseed rape and set-aside in the crop sequence decreased it. In winter oilseed rape, sequences with roots and tubers also increased fungicide use. In sugar beets, sequences with maize or a high concentration of sugar beets led to increasing fungicide use. If farmers chose riskier crop sequences tillage by plough decreased the need for herbicide and fungicide use. To reduce herbicide and fungicide use intensities we recommend increasing the diversity of crop rotations, including a higher number of crops per rotation together with ploughing. Simplifying both crop sequence diversity and tillage intensity implies higher use of herbicides and fungicides. Results will be useful for convincing farmers to diversify crop sequences.

Herbicide use simplified cropping systems, which lead to short rotations and sowing dates rather focusing on yield than on weed management. Re-diversifying these cropping systems should reduce the vast use of herbicides. We hypothesize that herbicide use decreases with more diverse cropping systems even with few common crops. Cropping system experiments raise methodological questions connected with herbicide use. So far, herbicide use in field experiments is a-priori fixed and not adapted to changing situations. To overcome this, we developed a heuristic decision rule to employ herbicide intensities as response variable in statistical models. Field trials were set up at two locations in Germany as split plot design with cropping systems as main factor and weed control targets as sub-factor. The cropping systems contained maize, winter wheat and winter oilseed rape, and differed in length and sowing date of winter wheat. Weed control targeted a high, medium and low multi-year level. The outcome ‘herbicide intensity’ was described as treatment frequency index (TFI) and as number of active ingredients (NAI). Finally, crops were harvested and yields measured. The heuristic decision rule enabled reproducible, objective herbicide decisions to adapt herbicide intensities to various weed infestations. Herbicide intensities sorted themselves according to the multi-year control target. All crops required a lower herbicide intensity when grown in longer rotations: on average TFI decreased by 19% and 11% resp. depending on the location, and NAI decreased by 0.3 on average per year. Highest yields were harvested with a medium to high weed control target. The heuristic approach opens up new opportunities to collect data on herbicide use as a dependent variable. More diversity in cropping systems reduced herbicide intensity, nevertheless, could not completely compensate the use of herbicides. Thus, we consider them as a useful tool in integrated weed management.

In scientific and public communities pesticide use in arable cropping is a strongly discussed management measure. Nevertheless, information about on-farm pesticide use intensities is scarce. We studied the contribution of regional and farm effects to average pesticide use and its variability and major influencing factors. Data on chemical plant protection measures of 60 farms in four regions of Northern Germany were examined for twelve field crops between 2005 and 2014. The regions correspond to administrative districts of similar size located on gradients of climatic, soil, and structural production conditions. First, we analysed pesticide use intensity described as Standardised Treatment Index (STI) per district and crop. Second, we estimated the variance components on STIs (‘District’, ‘Farm’, ‘Field’, ‘Year’) for total and groups of pesticides. We found a decreasing gradient in pesticide use intensity from the west to the east in Northern Germany. STIs were lower in districts with less favourable conditions for arable land use. However, analysis of variance components showed that the influence of the natural gradient is smaller than the influence of the farm. Hence, differences in farm management and socio-economic characteristics superimpose the production conditions. Finally, we checked the meaningfulness of our results by comparing them with previous and on-going large scale surveys. We determined differences (lower STIs) to the statistical means of the German reference farm network Plant Protection. We assume that this is caused by the higher aggregation level of regions of the German reference farm network Plant Protection.

Data on agricultural pesticide use are of increasing interest since understanding farmers’ pesticide use patterns is crucial for farm business, farm advisory, and agricultural and environmental policy. Member states of the European Union established different systems to estimate applied amounts of active ingredients per crop and hectare. However, the aggregation of use data at crop level lacks information of farm-specific use patterns. Former studies showed that pesticide use varies strongly between farms. Consequently, approaches are needed to obtain more farm or field-specific data of pesticide use. Glyphosate is the most applied active ingredient worldwide and widely used in Germany. Studies have been carried out in recent years to analyze glyphosate use patterns in German arable farming. For this study, data of a glyphosate-related survey conducted among German farmers were compared with field record data from on-farm studies. Using data originating from the two different sources will help to judge the comparability and adequacy of the methods. This comparison has not been done before for the case of glyphosate or any other pesticide. Questionnaire-based surveys among farmers correspond predominantly with field records, even in the highly political case of glyphosate. In view of glyphosate use patterns, the study indicates that many farmers are not fully aware of the difference between two uses currently registered: stubble and pre-sowing applications. In the on-going public debate, consistent definitions of glyphosate uses are needed to ensure clear communication between stakeholders.

We present an approach of Linear Mixed Effects Modelling to analyse the influence of crop management factors like rotation, tillage, seeding time and cultivar on pesticide use intensity from on-farm data. We use data from 761 winter wheat crops in eight commercial farms and six different regions of Eastern Germany. By defining cropping year, farm and region as random effects, the variability caused by these external factors can be accounted for before estimating effects of crop management factors. The objective of the research we present was to develop and test a method to analyse pesticide use monitoring data appropriately in order to (1) gain insight into the relationship between crop management and pesticide use, (2) integrate external sources of variability into the analysis and quantify their effect, and thereby (3) make a large data pool accessible for further interpretation which will develop further because this kind of monitoring has recently been made mandatory throughout Europe. The benefit of such an analysis is to utilise the practical experience generated by farmers with their daily work that is mirrored in these data for research purposes. This could be valuable in, for example, assessing the effect of Integrated Pest Management (IPM) measures on the pesticide use intensity in arable agriculture, for which evidence is still scarce. We quantified pesticide use intensity using the Standardised Treatment Index (STI). For fungicides, herbicides and growth regulators, we modelled the STI as response of 11 management variables using Linear Mixed Effects Modelling. Insecticide use intensity was not modelled since most fields were not treated with insecticides. Our results confirmed that pesticide use variability is mainly caused by external sources, but pesticide use intensity is significantly connected to crop management, mainly to preceding crop, seeding time and cultivar characteristics. Pesticide use was smaller when preventive measures typical for IPM were applied. We show that the Linear Mixed Effects Modelling approach with multiple predictor variables is well suited for analysing the connection between crop management and pesticide use. We discuss how the random variables influence pesticide use and treatment decisions. Our findings imply that future pesticide use analysis must take regional and farm specific factors into account more strongly, especially when deriving statistical means from on-farm samples. We conclude that the pesticide use monitoring could serve as a valuable data base for scientific research on IPM when data acquisition and analysis are well exploited. Results may be useful for convincing farmers to adopt cropping systems which are less dependent on pesticide use.

We analysed pesticide use data of 539 winter wheat fields grown on seven commercial farms in North East Germany in the seasons 2001–2007. The data were originally collected as part of economic analysis. The aim of our study was to show the impact of agronomic measures typical for integrated pest management (IPM) on pesticide use intensity. We used multivariate methods including hierarchical clustering and unconstrained ordination to describe variability of pesticide use and to identify pesticide use patterns. 166 fields had a complete record of eight crop management variables characterising rotation, tillage, cultivar, seeding time and N fertilisation. We subjected pesticide data of these fields to an ordination constrained by crop management data and considered weather and farm characteristics as co-variables influencing the pesticide use intensity. Crop management had a measurable effect even beyond the single farm and changing weather conditions: it accounted for approx. 10% of the explained variability of pesticide use. Pesticide intensity is adjusted by farmers to overall cropping intensity. Fields which were managed with typical IPM characteristics like late seeding and resistant cultivars were treated with a lower pesticide intensity. The downside of our findings is that these lower input CM systems are typically used only on less productive sites. Variability of pesticide use was mainly influenced by the farm, this factor accounted for 37% of explainable variance. We discuss two components of this high explanatory power: (I) farm characteristics which determine overall cropping intensity and (II) the routine of treatment decisions on the farm. We discuss the implications of our results regarding robust study design and sampling for pesticide use monitoring. To enhance IPM adoption in the context of recent European pesticide regulations, we identify the need for: (I) studies on the economy of reducing pesticide use by integrating agronomic measures, and (II) identification of innovative farmers for possible farmer to farmer exchange on successful IPM adoption.

Data on the occurrence of weed species in 1463 German oilseed rape (OSR) fields were used to multivariately investigate the interactions of management practices and ecological processes with weed species composition. Partial canonical correspondence analysis (pCCA) served to quantify the relative contribution of 25 characteristics of climate, site and crop to weed species composition. According to pCCA the considerable differences in weed species composition of OSR fields were mainly associated with the crop preceding OSR, the tillage intensity and soil quality. Longitude and precipitation were the most important environmental parameters. Factors driving the occurrence of weed species differed for the categories ‘common’, ‘frequent’ and ‘rare’ weed species. Common weed species showed the strongest response to factors acting at the field scale such as crop sowing date or soil quality, while only frequent species were affected by local environment, and the occurrence of rare weeds more than the other groups depended on large-scale geographical position variables as well as cropping intensity. Cluster analyses of thematically grouped explanatory variables resulted in the identification of four farming systems with different soil properties and management, two farming systems differing in the occurrence of OSR cropping, and five ecoregions, each with characteristic and statistically verified OSR weed communities.

In this study a new field of action for outcome-oriented payment schemes (payments by results) is introduced. That is the promotion of extensive summer grazing activities of livestock. Combining this objective with methods of outcome-oriented rewarding means to tie the payment to vegetation criteria indicating regular grazing activities instead of regularly controlling the presence of livestock on the field. In comparison to action-oriented concepts, the main advantage of this approach is that only one inspection walk is required. The paper describes and evaluates the first attempt to determine suitable indicator species for such a scheme. We pre-tested the approach with existing vegetation data only. A database of releves, plant species and agriculturally used grassland communities of the regional district Doberan, Mecklenburg-Western Pomerania (Germany) was utilised. Grazing activities can be identified by the existence of a specific number of vascular species indicating grazing activities on a particular grassland field. We utilised the indicator function of grassland communities for management conditions to resolve the deficiency of data base releves for providing no data about the management regime. The pasture community Lolio perennis-Cynosuretum cristati was used as a reference system to pre-test the suitability of different combinations of potential indicator species for separating pasture from non-pasture grassland. A tentative quantitative selection criterion for pasture grassland was proposed. That is the number of indicator species a releve/grassland field has to hold to be identified as pasture grassland. Database releves are not sufficient for calibrating such a concept, though. It is necessary to adjust the indicator species list to other drivers like use characteristics, nutrient availability, and soil moisture level. Here additional fieldwork is needed. As a first evidence for the validity of the concept, the suitability of the indicator list was measured by testing the correct identification of pasture grassland in a long-term grassland experiment built by the treatments “grazing” and “mowing”. ANOVA tests on the precision of the indicator species showed significant results. In the discussion a modification of the inventory method of grassland fields was recommended. The precision of the developed concept could be further enhanced by introducing the species dominance as a second factor deciding about the correct identification of pasture use.

In intensive arable areas, field margins can constitute boundary biotopes that are valuable for natural resource conservation. In order to analyse the potential for implementing more field margin measures, 865 responding farmers in intensive arable areas in three German federal states were surveyed. Determinants of their willingness to participate in field margin measures were analysed by binary logistic regression. The farmers were asked about the optimal design of field margins, and these designs are discussed within the context of nature conservation goals. The results show that farmers implement field margins less often if they are risk-averse. Farmers already implementing grassland environmental measures and mulch seeding or extensification measures with regards to herbicide and fertiliser reductions on arable land more often adopt field margin measures. Regarding the design of field margins, farmers’ preferences hardly differ between the federal states, and they do not significantly conflict with nature conservation needs in intensive arable areas. The main fear is the risk of weed spreading. Therefore, farmers prefer sown field margins, mown at least once a year and treated with restricted herbicide applications. Certain features, such as seed mixture, mowing date and plant protection measures, could be adapted in order to minimise conflicts with nature conservation objectives.

German plant protection law demands farmers to limit the use of pesticides to a necessary extent. In this article, we have developed a theoretical framework on deciding on pesticide use. We look at three different ways of interpreting how much pesticide is “necessary” to control a pest on a crop: (1) using a product at the recommended dose when a pest is found or a precautionary treatment thought necessary, (2) optimising pesticide use for economic savings through adjusted doses, thresholds, etc. and (3) minimising pesticide need by altering the cultivation system to lower the risk of pests. The smallest pesticide use intensity can be reached by combining approaches 2 and 3. In practice, information on optimising pesticide use needs further research to reduce the uncertainty about the effects of agronomical control measures, both with respect to economics and risk.

It is investigated, whether the arable weed vegetation can perform as a characteristic to stimulate sustainable development of agriculture in Central Europe. The concept of sustainable development represents a dynamic process, influenced by ecological, economic and social aspects. Characteristics are required to integrate desired ecological aspects into farming concepts, in order to stimulate production systems which add to a sustainable development of the sector. The investigations rely on data of an arable farming system experiment (INTEX). Three farming systems called good-farming-practice (GFP), integrated-flexible and integrated-non-inversion were established from 1994 to 1998 on two sites. The systems differed in crop rotation, soil tillage, fertilisation and pesticide use. Thus, weed management comprised cultural or indirect control by plant husbandry and direct mechanical and chemical control. A residual weed vegetation after terminating all direct control in the arable crops was ex-ante intended in the integrated systems. The spring weed densities were observed in winter wheat previous to any direct control treatments. The residual weed vegetation was investigated in all crops by monitoring species numbers and total ground cover after terminating all direct control. The direct chemical control intensity was considerably lower in the integrated farming system, in which the crop rotation was extended and the soil was ploughed annually. Spring densities of problematic weeds like annual grass weeds and Galium aparine in the winter wheat crop were not higher in these systems. Higher covers of the residual weed vegetation occurred more often in the productive arable crops of the two integrated farming systems, which additionally hosted higher number of species after terminating direct control. The different system strategies, including a higher input of herbicides in the system integrated-non-inversion, averted severe problems, including those arising from the dominating species being generally problematic in arable land use.

This paper focuses on the possibilities to reward ecological goods and services of agriculture. Ecological goods in agriculture are achieved by various types of agricultural land use either creating resources or buffering resource consumption. From an economic point of view, positive consequences will result from rewarding the agricultural sector for the production of ecological goods according to market principles. A rewarding system can be developed according to market principles when results instead of actions are rewarded, producer and consumer surpluses are created, demand can be expressed and supply can be provided. Ecological goods are public goods, therefore the expression of demand is up to the community. However, the purchasing of such ecological goods is up to the authority administrating public finances. Establishing regional markets is necessary, since regional differences are evident. Principally, the price of an ecological good should be determined by its value to the public, not by the cost of production. With respect to the various ecological goods achieved by the agricultural sector, the contribution of agricultural land use to biodiversity, and more specifically to plant biodiversity seems to be the most appropriate for a market-orientated rewarding system. A framework for establishing such a market is outlined in this paper. Deviation from standard economic principles regarding prices for ecological goods are considered to be necessary during an introductory period. Finally based on both, ecological and economical considerations, and regarding the intrinsic and extrinsic characteristics of “plant biodiversity connected with agricultural land use”, it seems worth the effort to further examine and develop this concept.

The determination of fixed guide values for weed economic threshold in cereals is briefly reviewed. A 2-year experimental series in winter cereals to validate these threshold values in the Federal Republic of Germany is described. It was found that the use of fixed threshold values to decide when to control annual weeds was more profitable than prophylactic herbicide application. The fixed threshold values are 20–30 plants m −2 for grass weeds, broad-leaved weeds 40–50 plants m −2 , Galium aparine 0.1–0.5 plants m −2 and Fallopia convolvulus 2.0 plants m −2 . Alternatively, broad-leaved weeds have to be assessed as percentage ground cover, 5–10% being the threshold. In order to improve further the predictions from the economic thresholds a more flexible computer-based decision model has been developed. This includes updated biological data and simple economic calculations. Further experiments have been done to validate the model. Assessments based on the calculated net return indicate that weed control decisions were slightly improved by model recommendations compared with the fixed threshold concept. The basic assumptions in the decision model are discussed and possibilities for future applications under practical conditions are described.